On Thu Aug 28, 2025 at 7:18 AM +08, Andrii Nakryiko wrote:
> On Wed, Aug 27, 2025 at 9:46 AM Leon Hwang <leon.hwang@xxxxxxxxx> wrote:
>>
>> Add libbpf support for the BPF_F_CPU flag for percpu maps by embedding the
>> cpu info into the high 32 bits of:
>>
>> 1. **flags**: bpf_map_lookup_elem_flags(), bpf_map__lookup_elem(),
>> bpf_map_update_elem() and bpf_map__update_elem()
>> 2. **opts->elem_flags**: bpf_map_lookup_batch() and
>> bpf_map_update_batch()
>>
>> And the flag can be BPF_F_ALL_CPUS, but cannot be
>> 'BPF_F_CPU | BPF_F_ALL_CPUS'.
>>
>> Behavior:
>>
>> * If the flag is BPF_F_ALL_CPUS, the update is applied across all CPUs.
>> * If the flag is BPF_F_CPU, it updates value only to the specified CPU.
>> * If the flag is BPF_F_CPU, lookup value only from the specified CPU.
>> * lookup does not support BPF_F_ALL_CPUS.
>>
>> Signed-off-by: Leon Hwang <leon.hwang@xxxxxxxxx>
>> ---
>> tools/lib/bpf/bpf.h | 8 ++++++++
>> tools/lib/bpf/libbpf.c | 33 +++++++++++++++++++++++++++------
>> tools/lib/bpf/libbpf.h | 21 ++++++++-------------
>> 3 files changed, 43 insertions(+), 19 deletions(-)
>>
>
> [...]
>
>> @@ -10629,6 +10629,27 @@ static int validate_map_op(const struct bpf_map *map, size_t key_sz,
>> case BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE: {
>> int num_cpu = libbpf_num_possible_cpus();
>> size_t elem_sz = roundup(map->def.value_size, 8);
>> + __u32 cpu;
>> +
>> + if (flags & (BPF_F_CPU | BPF_F_ALL_CPUS)) {
>> + if ((flags & BPF_F_CPU) && (flags & BPF_F_ALL_CPUS)) {
>> + pr_warn("map '%s': can't use BPF_F_CPU and BPF_F_ALL_CPUS at the same time\n",
>> + map->name);
>> + return -EINVAL;
>> + }
>> + cpu = flags >> 32;
>> + if (cpu >= num_cpu) {
>
> only check this if BPF_F_CPU is set
>
>> + pr_warn("map '%s': cpu %u in flags cannot be GE num cpus %d\n",
>
> "GE"? maybe "CPU #%d is not valid"... (or don't even check cpu value
> itself, it's unlikely user using BPF_F_CPU will get it so wrong)
>
Let's remove it instead.
Thanks,
Leon
