Re: [PATCH bpf-next v9 3/9] libbpf: Probe percpu data feature

[Leon Hwang <leon.hwang@linux.dev>]

On 14/7/26 00:17, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] The capability probe for percpu data incorrectly emits warnings and returns a negative error code when map creation fails on older kernels, breaking the silent probing expectation and causing double warnings.
> --
> 
> --- Patch [3]: [PATCH bpf-next v9 3/9] libbpf: Probe percpu data feature ---
> commit 4bf6930d4ace160491d05ec9880c9cadc9f81bc7
> Author: Leon Hwang <leon.hwang@linux.dev>
> 
> libbpf: Probe percpu data feature
> 
> This commit adds a dedicated feature probe to libbpf for detecting whether
> the kernel supports global percpu data. It checks this capability by
> attempting to create a simple percpu array map and loading a basic socket
> filter program that accesses it.
> 
>> diff --git a/tools/lib/bpf/features.c b/tools/lib/bpf/features.c
>> index b7e388f99d0bb..ef9581c113035 100644
>> --- a/tools/lib/bpf/features.c
>> +++ b/tools/lib/bpf/features.c
> 
> [ ... ]
> 
>> @@ -620,6 +620,38 @@ static int probe_bpf_syscall_common_attrs(int token_fd)
> 
> [ ... ]
> 
>> +static int probe_kern_percpu_data(int token_fd)
>> +{
> 
> [ ... ]
> 
>> +	map = bpf_map_create(BPF_MAP_TYPE_PERCPU_ARRAY, "libbpf_percpu", sizeof(int), 8, 1,
>> +			     &map_opts);
>> +	if (map < 0) {
>> +		pr_warn("Error in %s(): %s. Couldn't create simple percpu_array map.\n",
>> +			__func__, errstr(map));
>> +		return map;
>> +	}
> 
> [Severity: Medium]
> Should this probe silently return 0 instead of propagating the negative error
> code and printing a warning when bpf_map_create() fails?
> 
> When running on older kernels that genuinely lack BPF_MAP_TYPE_PERCPU_ARRAY
> support, this code will print the pr_warn() above and return a negative error
> code.
> 
> Because the caller feat_supported() expects 0 to cleanly indicate a missing
> feature, returning a negative value causes feat_supported() to print a second
> warning: "Detection of kernel kernel supports percpu data support failed".
> 
> Does this unintentionally pollute the logs with double warnings when apps are
> safely falling back on older kernels?
> 


Cite my reply in v5 [1]:

Since percpu_array map is a basic map in bpf, this map creation should
likely succeed. If it fails, the kernels are too old to use such new libbpf.

[1]
https://lore.kernel.org/bpf/66d40066-94a7-4710-8e0f-d18f8c1bf21f@linux.dev/

Thanks,
Leon