Re: [PATCH bpf-next v6 10/12] selftests/bpf: Add tests to verify verifier log for global percpu data

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] Incorrect `data_len` argument passed to `bpf_snprintf` makes the negative test brittle.
--

commit d5b0b79b30e924d1fec11028be9289d47ccc5221
Author: Leon Hwang <leon.hwang@linux.dev>

selftests/bpf: Add tests to verify verifier log for global percpu data

This commit adds new eBPF programs to verify the verifier log correctly
rejects bpf_strncmp and bpf_snprintf when a percpu array map is used as a
const string. It also updates the prog test to execute these new checks.

> diff --git a/tools/testing/selftests/bpf/progs/test_global_percpu_data.c b/tools/testing/selftests/bpf/progs/test_global_percpu_data.c
> index ba92ffb0ca499..2222ad3a49bd7 100644
> --- a/tools/testing/selftests/bpf/progs/test_global_percpu_data.c
> +++ b/tools/testing/selftests/bpf/progs/test_global_percpu_data.c

[ ... ]

> +SEC("?kprobe")
> +__failure __msg("R{{[0-9]+}} points to percpu_array map which cannot be used as const string")
> +int verifier_snprintf(void *ctx)
> +{
> +	u64 args[] = { data };
> +	char buf[128];
> +	int len;
> +
> +	len = bpf_snprintf(buf, sizeof(buf), fmt, args, 1);

[Severity: Low]
Should this be passing sizeof(args) instead of 1?

The bpf_snprintf helper requires data_len to be the size of the data array
in bytes. Passing 1 makes the test brittle if the verifier's argument
evaluation order changes in the future, as it would fail on the data_len
argument before checking the fmt argument.

> +	if (len > 0)
> +		bpf_printk("snprintf: %s\n", buf);
> +	return 0;
> +}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260615152646.27639-1-leon.hwang@linux.dev?part=10