Re: [PATCH v3] cgroup/bpf: fast path skb BPF filtering

[Pavel Begunkov <asml.silence@gmail.com>]

On 12/15/21 16:51, sdf@google.com wrote:
> On 12/15, Pavel Begunkov wrote:
>> Add per socket fast path for not enabled BPF skb filtering, which sheds
>> a nice chunk of send/recv overhead when affected. Testing udp with 128
>> byte payload and/or zerocopy with any payload size showed 2-3%
>> improvement in requests/s on the tx side using fast NICs across network,
>> and around 4% for dummy device. Same goes for rx, not measured, but
>> numbers should be relatable.
>> In my understanding, this should affect a good share of machines, and at
>> least it includes my laptops and some checked servers.
> 
>> The core of the problem is that even though there is
>> cgroup_bpf_enabled_key guarding from __cgroup_bpf_run_filter_skb()
>> overhead, there are cases where we have several cgroups and loading a
>> BPF program to one also makes all others to go through the slow path
>> even when they don't have any BPF attached. It's even worse, because
>> apparently systemd or some other early init loads some BPF and so
>> triggers exactly this situation for normal networking.
> 
>> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
>> ---
> 
>> v2: replace bitmask appoach with empty_prog_array (suggested by Martin)
>> v3: add "bpf_" prefix to empty_prog_array (Martin)
> 
>>   include/linux/bpf-cgroup.h | 24 +++++++++++++++++++++---
>>   include/linux/bpf.h        | 13 +++++++++++++
>>   kernel/bpf/cgroup.c        | 18 ++----------------
>>   kernel/bpf/core.c          | 16 ++++------------
>>   4 files changed, 40 insertions(+), 31 deletions(-)
> 
>> diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h
>> index 11820a430d6c..c6dacdbdf565 100644
>> --- a/include/linux/bpf-cgroup.h
>> +++ b/include/linux/bpf-cgroup.h
>> @@ -219,11 +219,28 @@ int bpf_percpu_cgroup_storage_copy(struct bpf_map *map, void *key, void *value);
>>   int bpf_percpu_cgroup_storage_update(struct bpf_map *map, void *key,
>>                        void *value, u64 flags);
> 
>> +static inline bool
>> +__cgroup_bpf_prog_array_is_empty(struct cgroup_bpf *cgrp_bpf,
>> +                 enum cgroup_bpf_attach_type type)
>> +{
>> +    struct bpf_prog_array *array = rcu_access_pointer(cgrp_bpf->effective[type]);
>> +
>> +    return array == &bpf_empty_prog_array.hdr;
>> +}
>> +
>> +#define CGROUP_BPF_TYPE_ENABLED(sk, atype)                       \
>> +({                                           \
>> +    struct cgroup *__cgrp = sock_cgroup_ptr(&(sk)->sk_cgrp_data);           \
>> +                                           \
>> +    !__cgroup_bpf_prog_array_is_empty(&__cgrp->bpf, (atype));           \
>> +})
>> +
>>   /* Wrappers for __cgroup_bpf_run_filter_skb() guarded by cgroup_bpf_enabled. */
>>   #define BPF_CGROUP_RUN_PROG_INET_INGRESS(sk, skb)                  \
>>   ({                                          \
>>       int __ret = 0;                                  \
>> -    if (cgroup_bpf_enabled(CGROUP_INET_INGRESS))              \
>> +    if (cgroup_bpf_enabled(CGROUP_INET_INGRESS) && sk &&              \
>> +        CGROUP_BPF_TYPE_ENABLED((sk), CGROUP_INET_INGRESS))           \
> 
> Why not add this __cgroup_bpf_run_filter_skb check to
> __cgroup_bpf_run_filter_skb? Result of sock_cgroup_ptr() is already there
> and you can use it. Maybe move the things around if you want
> it to happen earlier.

For inlining. Just wanted to get it done right, otherwise I'll likely be
returning to it back in a few months complaining that I see measurable
overhead from the function call :)

-- 
Pavel Begunkov