netdev - Re: [PATCH v3 bpf-next 1/2] bpf: Extend batch operations for map-in-map bpf-maps

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAEf4BzZ6oTz=gojxFvPKnrJ_cV_eS_Ra7YZDiJvfTdteEC8d2w@mail.gmail.com>
Date:   Wed, 27 Apr 2022 13:15:46 -0700
From:   Andrii Nakryiko <andrii.nakryiko@...il.com>
To:     Takshak Chahande <ctakshak@...com>, Yonghong Song <yhs@...com>
Cc:     Networking <netdev@...r.kernel.org>, bpf <bpf@...r.kernel.org>,
        Andrii Nakryiko <andrii@...nel.org>,
        Alexei Starovoitov <ast@...nel.org>, ndixit@...com,
        Martin Lau <kafai@...com>, Andrii Nakryiko <andriin@...com>,
        Daniel Borkmann <daniel@...earbox.net>
Subject: Re: [PATCH v3 bpf-next 1/2] bpf: Extend batch operations for
 map-in-map bpf-maps

On Mon, Apr 25, 2022 at 11:42 AM Takshak Chahande <ctakshak@...com> wrote:
>
> This patch extends batch operations support for map-in-map map-types:
> BPF_MAP_TYPE_HASH_OF_MAPS and BPF_MAP_TYPE_ARRAY_OF_MAPS
>
> A usecase where outer HASH map holds hundred of VIP entries and its
> associated reuse-ports per VIP stored in REUSEPORT_SOCKARRAY type
> inner map, needs to do batch operation for performance gain.
>
> This patch leverages the exiting generic functions for most of the batch
> operations. As map-in-map's value contains the actual reference of the inner map,
> for BPF_MAP_TYPE_HASH_OF_MAPS type, it needed an extra step to fetch the
> map_id from the reference value.
>
> selftests are added in next patch that has v1->v3 changes
>
> Signed-off-by: Takshak Chahande <ctakshak@...com>
> ---

cc'ing Yonghong who was involved in designing and implementing these
batch APIs. PTAL when you get a chance, thanks!

>  kernel/bpf/arraymap.c |  2 ++
>  kernel/bpf/hashtab.c  | 12 ++++++++++--
>  2 files changed, 12 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c
> index 7f145aefbff8..f0852b6617cc 100644
> --- a/kernel/bpf/arraymap.c
> +++ b/kernel/bpf/arraymap.c
> @@ -1344,6 +1344,8 @@ const struct bpf_map_ops array_of_maps_map_ops = {
>         .map_fd_put_ptr = bpf_map_fd_put_ptr,
>         .map_fd_sys_lookup_elem = bpf_map_fd_sys_lookup_elem,
>         .map_gen_lookup = array_of_map_gen_lookup,
> +       .map_lookup_batch = generic_map_lookup_batch,
> +       .map_update_batch = generic_map_update_batch,
>         .map_check_btf = map_check_no_btf,
>         .map_btf_name = "bpf_array",
>         .map_btf_id = &array_of_maps_map_btf_id,
> diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
> index c68fbebc8c00..fd537bfba84c 100644
> --- a/kernel/bpf/hashtab.c
> +++ b/kernel/bpf/hashtab.c
> @@ -139,7 +139,7 @@ static inline bool htab_use_raw_lock(const struct bpf_htab *htab)
>
>  static void htab_init_buckets(struct bpf_htab *htab)
>  {
> -       unsigned i;
> +       unsigned int i;
>
>         for (i = 0; i < htab->n_buckets; i++) {
>                 INIT_HLIST_NULLS_HEAD(&htab->buckets[i].head, i);
> @@ -1594,7 +1594,7 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map,
>         void __user *uvalues = u64_to_user_ptr(attr->batch.values);
>         void __user *ukeys = u64_to_user_ptr(attr->batch.keys);
>         void __user *ubatch = u64_to_user_ptr(attr->batch.in_batch);
> -       u32 batch, max_count, size, bucket_size;
> +       u32 batch, max_count, size, bucket_size, map_id;
>         struct htab_elem *node_to_free = NULL;
>         u64 elem_map_flags, map_flags;
>         struct hlist_nulls_head *head;
> @@ -1719,6 +1719,13 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *map,
>                         }
>                 } else {
>                         value = l->key + roundup_key_size;
> +                       if (map->map_type == BPF_MAP_TYPE_HASH_OF_MAPS) {
> +                               struct bpf_map **inner_map = value;
> +                                /* Actual value is the id of the inner map */
> +                               map_id = map->ops->map_fd_sys_lookup_elem(*inner_map);
> +                               value = &map_id;
> +                       }
> +
>                         if (elem_map_flags & BPF_F_LOCK)
>                                 copy_map_value_locked(map, dst_val, value,
>                                                       true);
> @@ -2425,6 +2432,7 @@ const struct bpf_map_ops htab_of_maps_map_ops = {
>         .map_fd_sys_lookup_elem = bpf_map_fd_sys_lookup_elem,
>         .map_gen_lookup = htab_of_map_gen_lookup,
>         .map_check_btf = map_check_no_btf,
> +       BATCH_OPS(htab),
>         .map_btf_name = "bpf_htab",
>         .map_btf_id = &htab_of_maps_map_btf_id,
>  };
> --
> 2.30.2
>