Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.31 as permitted sender) client-ip=23.128.96.31;
Message-ID: <0a142924-633c-44e6-9a92-2dc019656bf2@linux.dev>
Date:   Mon, 23 Oct 2023 20:28:21 -0700
MIME-Version: 1.0
Subject: Re: [PATCH] selftests: bpf: add malloc failures checks in bpf_iter
Content-Language: en-GB
To:     Yuran Pereira <yuran.pereira@hotmail.com>, shuah@kernel.org
Cc:     ast@kernel.org, daniel@iogearbox.net, andrii@kernel.org,
        martin.lau@linux.dev, song@kernel.org, john.fastabend@gmail.com,
        kpsingh@kernel.org, sdf@google.com, haoluo@google.com,
        jolsa@kernel.org, mykolal@fb.com, brauner@kernel.org,
        iii@linux.ibm.com, kuifeng@meta.com, bpf@vger.kernel.org,
        linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org
References: <DB3PR10MB683506E8DCCB073A2B75BB10E8DFA@DB3PR10MB6835.EURPRD10.PROD.OUTLOOK.COM>
From:   Yonghong Song <yonghong.song@linux.dev>
In-Reply-To: <DB3PR10MB683506E8DCCB073A2B75BB10E8DFA@DB3PR10MB6835.EURPRD10.PROD.OUTLOOK.COM>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk


On 10/23/23 7:59 PM, Yuran Pereira wrote:
> Since some malloc calls in bpf_iter may at times fail,
> this patch adds the appropriate fail checks, and ensures that
> any previously allocated resource is appropriately destroyed
> before returning the function.
>
> Signed-off-by: Yuran Pereira <yuran.pereira@hotmail.com>
> ---
>   tools/testing/selftests/bpf/prog_tests/bpf_iter.c | 10 ++++++++++
>   1 file changed, 10 insertions(+)
>
> diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
> index 1f02168103dd..6d47ea9211a4 100644
> --- a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
> +++ b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
> @@ -878,6 +878,11 @@ static void test_bpf_percpu_hash_map(void)
>   
>   	skel->rodata->num_cpus = bpf_num_possible_cpus();
>   	val = malloc(8 * bpf_num_possible_cpus());
> +	if (CHECK(!val, "malloc", "memory allocation failed: %s",
> +				strerror(errno))) {
> +		bpf_iter_bpf_percpu_hash_map__destroy(skel);
> +		return;
> +	}


Could you use !ASSERT_OK_PTR(...) instead of CHECK? bpf selftest
prefers to use ASSERT_* series of macros instead of CHECK.
In the above example, printing out strerror is not required,
see some other examples in the same folder.

Also bpf_iter.c has some other usages of CHECK macro.
Since you are touching this file, could you convert all
CHECK macros to ASSERT_* macros?

Basically you need two patches:
   patch 1: convert existing CHECK macros to ASSERT_* macros in bpf_iter.c.
            this should not check any functionality except error messages.
   patch 2: your patch with ASSERT_* macros.

You can use the following as your subject line:
   [PATCH bpf-next] selftests/bpf: Add malloc failure checks in bpf_iter

>   
>   	err = bpf_iter_bpf_percpu_hash_map__load(skel);
>   	if (!ASSERT_OK_PTR(skel, "bpf_iter_bpf_percpu_hash_map__load"))
> @@ -1057,6 +1062,11 @@ static void test_bpf_percpu_array_map(void)
>   
>   	skel->rodata->num_cpus = bpf_num_possible_cpus();
>   	val = malloc(8 * bpf_num_possible_cpus());
> +	if (CHECK(!val, "malloc", "memory allocation failed: %s",
> +				strerror(errno))) {
> +		bpf_iter_bpf_percpu_hash_map__destroy(skel);
> +		return;
> +	}
>   
>   	err = bpf_iter_bpf_percpu_array_map__load(skel);
>   	if (!ASSERT_OK_PTR(skel, "bpf_iter_bpf_percpu_array_map__load"))