netdev - Re: [PATCH bpf v4 13/14] bpf/tests: Fix error in tail call limit tests

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

Message-ID: <4b9db215-edcd-6089-6ecd-6fe9b20dcbbb@loongson.cn>
Date:   Tue, 14 Sep 2021 20:41:48 +0800
From:   Tiezhu Yang <yangtiezhu@...ngson.cn>
To:     Johan Almbladh <johan.almbladh@...finetworks.com>, ast@...nel.org,
        daniel@...earbox.net, andrii@...nel.org
Cc:     kafai@...com, songliubraving@...com, yhs@...com,
        john.fastabend@...il.com, kpsingh@...nel.org, iii@...ux.ibm.com,
        paul@...ium.io, netdev@...r.kernel.org, bpf@...r.kernel.org
Subject: Re: [PATCH bpf v4 13/14] bpf/tests: Fix error in tail call limit
 tests

On 09/14/2021 05:18 PM, Johan Almbladh wrote:
> This patch fixes an error in the tail call limit test that caused the
> test to fail on for x86-64 JIT. Previously, the register R0 was used to
> report the total number of tail calls made. However, after a tail call
> fall-through, the value of the R0 register is undefined. Now, all tail
> call error path tests instead use context state to store the count.
>
> Fixes: 874be05f525e ("bpf, tests: Add tail call test suite")
> Reported-by: Paul Chaignon <paul@...ium.io>
> Reported-by: Tiezhu Yang <yangtiezhu@...ngson.cn>
> Signed-off-by: Johan Almbladh <johan.almbladh@...finetworks.com>
> ---
>   lib/test_bpf.c | 37 +++++++++++++++++++++++++++----------
>   1 file changed, 27 insertions(+), 10 deletions(-)
>
> diff --git a/lib/test_bpf.c b/lib/test_bpf.c
> index 7475abfd2186..ddb9a8089d2e 100644
> --- a/lib/test_bpf.c
> +++ b/lib/test_bpf.c
> @@ -12179,10 +12179,15 @@ static __init int test_bpf(void)
>   struct tail_call_test {
>   	const char *descr;
>   	struct bpf_insn insns[MAX_INSNS];
> +	int flags;
>   	int result;
>   	int stack_depth;
>   };
>   
> +/* Flags that can be passed to tail call test cases */
> +#define FLAG_NEED_STATE		BIT(0)
> +#define FLAG_RESULT_IN_STATE	BIT(1)
> +
>   /*
>    * Magic marker used in test snippets for tail calls below.
>    * BPF_LD/MOV to R2 and R2 with this immediate value is replaced
> @@ -12252,32 +12257,38 @@ static struct tail_call_test tail_call_tests[] = {
>   	{
>   		"Tail call error path, max count reached",
>   		.insns = {
> -			BPF_ALU64_IMM(BPF_ADD, R1, 1),
> -			BPF_ALU64_REG(BPF_MOV, R0, R1),
> +			BPF_LDX_MEM(BPF_W, R2, R1, 0),
> +			BPF_ALU64_IMM(BPF_ADD, R2, 1),
> +			BPF_STX_MEM(BPF_W, R1, R2, 0),
>   			TAIL_CALL(0),
>   			BPF_EXIT_INSN(),
>   		},
> -		.result = MAX_TAIL_CALL_CNT + 1,
> +		.flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE,
> +		.result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS,

Hi Johan,

I have tested this patch,
It should be "MAX_TAIL_CALL_CNT + 1" instead of "MAX_TAIL_CALL_CNT + 1 + 1"?

[...]

Thanks,
Tiezhu