From nobody Mon Dec 1 22:41:53 2025 Received: from out-183.mta0.migadu.com (out-183.mta0.migadu.com [91.218.175.183]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 67362304994; Wed, 26 Nov 2025 14:51:39 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.183 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168701; cv=none; b=sKl49zBDQ31GPqcuWtiZFZVf6YbPJ5WvHXBZsvxGbcVfDv7zD0YS70q7q+63UWlzP/s5aHUWc07uGUFJemXb9bJeRH9WsrqmyOPt7yOKtIYUjLSuDW//9Z1DIzMfadWsVev6CvFmW54jPmaFZm0SxtRW5AOwygLzRXXYxkp97o8= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168701; c=relaxed/simple; bh=z+aVWh1DG3zW7uBJWklHailp5mnhLIZWDprqOY0jiFk=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=fJ2VJVP1+hVXiUDXNRYkvXk80jJdcpKySjtzcQ/id3HTyJUXKh1TP7Afg3bPuSP2bKZbYkemfIqKVrjIUjSTgkDN9nP7Ra6EK68RP1+1KdZoqurKuZGZgvXkHc/qbhVbq3Wyfg8+YI/ZkllK+quid8Ki/wXMng+OngFWWi671j0= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=GZ/kgA1u; arc=none smtp.client-ip=91.218.175.183 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="GZ/kgA1u" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168696; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=j8/i3V0qKATfpWYij/aZrmcs/5FGPNFp1K/yt6QF2yc=; b=GZ/kgA1uDl/h/hqgaFqkf8h7noPRVBo0QscOPmTCq8Iw4jiGW5KR+yEwxxWguwsFWC1fXI 5b8Uy3WDnmsG4JZRaGMGM9Nyq6fy7GuHrXcShv4Ul2gnEZDeNR3ojwCR4uK8K4MmLbcOvN waRc6VK/9w02xqtv5WMY6VRVxsYvXhA= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 1/7] bpf: Introduce BPF_F_CPU and BPF_F_ALL_CPUS flags Date: Wed, 26 Nov 2025 22:50:33 +0800 Message-ID: <20251126145039.15715-2-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Introduce BPF_F_CPU and BPF_F_ALL_CPUS flags and check them for following APIs: * 'map_lookup_elem()' * 'map_update_elem()' * 'generic_map_lookup_batch()' * 'generic_map_update_batch()' And, get the correct value size for these APIs. Acked-by: Andrii Nakryiko Signed-off-by: Leon Hwang --- include/linux/bpf.h | 23 ++++++++++++++++++++++- include/uapi/linux/bpf.h | 2 ++ kernel/bpf/syscall.c | 31 +++++++++++++++++-------------- tools/include/uapi/linux/bpf.h | 2 ++ 4 files changed, 43 insertions(+), 15 deletions(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 6498be4c44f8..d84af3719b59 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -3829,14 +3829,35 @@ bpf_prog_update_insn_ptrs(struct bpf_prog *prog, u3= 2 *offsets, void *image) } #endif =20 +static inline bool bpf_map_supports_cpu_flags(enum bpf_map_type map_type) +{ + return false; +} + static inline int bpf_map_check_op_flags(struct bpf_map *map, u64 flags, u= 64 allowed_flags) { - if (flags & ~allowed_flags) + u32 cpu; + + if ((u32)flags & ~allowed_flags) return -EINVAL; =20 if ((flags & BPF_F_LOCK) && !btf_record_has_field(map->record, BPF_SPIN_L= OCK)) return -EINVAL; =20 + if (!(flags & BPF_F_CPU) && flags >> 32) + return -EINVAL; + + if (flags & (BPF_F_CPU | BPF_F_ALL_CPUS)) { + if (!bpf_map_supports_cpu_flags(map->map_type)) + return -EINVAL; + if ((flags & BPF_F_CPU) && (flags & BPF_F_ALL_CPUS)) + return -EINVAL; + + cpu =3D flags >> 32; + if ((flags & BPF_F_CPU) && cpu >=3D num_possible_cpus()) + return -ERANGE; + } + return 0; } =20 diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index f5713f59ac10..8b6279ca6e66 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -1373,6 +1373,8 @@ enum { BPF_NOEXIST =3D 1, /* create new element if it didn't exist */ BPF_EXIST =3D 2, /* update existing element */ BPF_F_LOCK =3D 4, /* spin_lock-ed map_lookup/map_update */ + BPF_F_CPU =3D 8, /* cpu flag for percpu maps, upper 32-bit of flags is a = cpu number */ + BPF_F_ALL_CPUS =3D 16, /* update value across all CPUs for percpu maps */ }; =20 /* flags for BPF_MAP_CREATE command */ diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index cef8963d69f9..7a1d0ce8fa57 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -133,12 +133,14 @@ bool bpf_map_write_active(const struct bpf_map *map) return atomic64_read(&map->writecnt) !=3D 0; } =20 -static u32 bpf_map_value_size(const struct bpf_map *map) -{ - if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_HASH || - map->map_type =3D=3D BPF_MAP_TYPE_LRU_PERCPU_HASH || - map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_ARRAY || - map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE) +static u32 bpf_map_value_size(const struct bpf_map *map, u64 flags) +{ + if (flags & (BPF_F_CPU | BPF_F_ALL_CPUS)) + return map->value_size; + else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_HASH || + map->map_type =3D=3D BPF_MAP_TYPE_LRU_PERCPU_HASH || + map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_ARRAY || + map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE) return round_up(map->value_size, 8) * num_possible_cpus(); else if (IS_FD_MAP(map)) return sizeof(u32); @@ -1732,7 +1734,7 @@ static int map_lookup_elem(union bpf_attr *attr) if (!(map_get_sys_perms(map, f) & FMODE_CAN_READ)) return -EPERM; =20 - err =3D bpf_map_check_op_flags(map, attr->flags, BPF_F_LOCK); + err =3D bpf_map_check_op_flags(map, attr->flags, BPF_F_LOCK | BPF_F_CPU); if (err) return err; =20 @@ -1740,7 +1742,7 @@ static int map_lookup_elem(union bpf_attr *attr) if (IS_ERR(key)) return PTR_ERR(key); =20 - value_size =3D bpf_map_value_size(map); + value_size =3D bpf_map_value_size(map, attr->flags); =20 err =3D -ENOMEM; value =3D kvmalloc(value_size, GFP_USER | __GFP_NOWARN); @@ -1807,7 +1809,7 @@ static int map_update_elem(union bpf_attr *attr, bpfp= tr_t uattr) goto err_put; } =20 - value_size =3D bpf_map_value_size(map); + value_size =3D bpf_map_value_size(map, attr->flags); value =3D kvmemdup_bpfptr(uvalue, value_size); if (IS_ERR(value)) { err =3D PTR_ERR(value); @@ -2003,11 +2005,12 @@ int generic_map_update_batch(struct bpf_map *map, s= truct file *map_file, void *key, *value; int err =3D 0; =20 - err =3D bpf_map_check_op_flags(map, attr->batch.elem_flags, BPF_F_LOCK); + err =3D bpf_map_check_op_flags(map, attr->batch.elem_flags, + BPF_F_LOCK | BPF_F_CPU | BPF_F_ALL_CPUS); if (err) return err; =20 - value_size =3D bpf_map_value_size(map); + value_size =3D bpf_map_value_size(map, attr->batch.elem_flags); =20 max_count =3D attr->batch.count; if (!max_count) @@ -2062,11 +2065,11 @@ int generic_map_lookup_batch(struct bpf_map *map, u32 value_size, cp, max_count; int err; =20 - err =3D bpf_map_check_op_flags(map, attr->batch.elem_flags, BPF_F_LOCK); + err =3D bpf_map_check_op_flags(map, attr->batch.elem_flags, BPF_F_LOCK | = BPF_F_CPU); if (err) return err; =20 - value_size =3D bpf_map_value_size(map); + value_size =3D bpf_map_value_size(map, attr->batch.elem_flags); =20 max_count =3D attr->batch.count; if (!max_count) @@ -2188,7 +2191,7 @@ static int map_lookup_and_delete_elem(union bpf_attr = *attr) goto err_put; } =20 - value_size =3D bpf_map_value_size(map); + value_size =3D bpf_map_value_size(map, 0); =20 err =3D -ENOMEM; value =3D kvmalloc(value_size, GFP_USER | __GFP_NOWARN); diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index f5713f59ac10..8b6279ca6e66 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -1373,6 +1373,8 @@ enum { BPF_NOEXIST =3D 1, /* create new element if it didn't exist */ BPF_EXIST =3D 2, /* update existing element */ BPF_F_LOCK =3D 4, /* spin_lock-ed map_lookup/map_update */ + BPF_F_CPU =3D 8, /* cpu flag for percpu maps, upper 32-bit of flags is a = cpu number */ + BPF_F_ALL_CPUS =3D 16, /* update value across all CPUs for percpu maps */ }; =20 /* flags for BPF_MAP_CREATE command */ --=20 2.51.2 From nobody Mon Dec 1 22:41:53 2025 Received: from out-178.mta0.migadu.com (out-178.mta0.migadu.com [91.218.175.178]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C99093081DF for ; Wed, 26 Nov 2025 14:51:48 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.178 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168710; cv=none; b=cLsFdeepQfePsbVr9dRzMoaDy608/EbaTkBrxcAHxq5znGcq1rd6uIaMs5PRPQMWDkpm/OAyc1WSD1YognsuoaIT78W0fjG3fagr4max1P/hpmE6VUavkAtbJyY0/Ceh00PGgWEPYx1MvkI4ifLsqRvfGryeJNQrT+rCDe8Vu30= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168710; c=relaxed/simple; bh=vE/JPMmNA+Dn2lRE547H5bvN6VnYzpVVr1Dc/xgSluc=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=cJs+Ske1bYIYswQRlpqxWeOtb4s5skeN4uMP0tLfZoU5DEcDrstDHL7k9BrZtkkzO/EWtEoiBs02SvQvoAk7IhP9M0TxuV7xhjaViws1TpE2P/rXfDDC5iehF/AJ+9k4BbteTQVTwMOXPr6S+8OAWP977KgyGSDZ4w/673ZldRQ= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=U70W6NrU; arc=none smtp.client-ip=91.218.175.178 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="U70W6NrU" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168706; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=BizZbmxMylmDygcgp8y5D9NypN6p1cqZpSSMPgibyaE=; b=U70W6NrULrKFMH/7K2KoBcnZFFWgfNzxdobhqHPd1Dg1QKnj/ws38pJ9iS0/hN1ujE2kDT mOevFF3ZCDQE7828jw9NWECp7LYYM9pBi7kdACZSKVk4TKewnf9d4OtaaHyECgm4jteJEO Of3X7yUpZmHNRwJxhzHheIxvq2ceX4Y= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 2/7] bpf: Add BPF_F_CPU and BPF_F_ALL_CPUS flags support for percpu_array maps Date: Wed, 26 Nov 2025 22:50:34 +0800 Message-ID: <20251126145039.15715-3-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Introduce support for the BPF_F_ALL_CPUS flag in percpu_array maps to allow updating values for all CPUs with a single value for both update_elem and update_batch APIs. Introduce support for the BPF_F_CPU flag in percpu_array maps to allow: * update value for specified CPU for both update_elem and update_batch APIs. * lookup value for specified CPU for both lookup_elem and lookup_batch APIs. The BPF_F_CPU flag is passed via: * map_flags of lookup_elem and update_elem APIs along with embedded cpu info. * elem_flags of lookup_batch and update_batch APIs along with embedded cpu info. Signed-off-by: Leon Hwang --- include/linux/bpf.h | 9 +++++++-- kernel/bpf/arraymap.c | 29 +++++++++++++++++++++++------ kernel/bpf/syscall.c | 2 +- 3 files changed, 31 insertions(+), 9 deletions(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index d84af3719b59..01a99e3a3e51 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -2762,7 +2762,7 @@ int map_set_for_each_callback_args(struct bpf_verifie= r_env *env, struct bpf_func_state *callee); =20 int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value); -int bpf_percpu_array_copy(struct bpf_map *map, void *key, void *value); +int bpf_percpu_array_copy(struct bpf_map *map, void *key, void *value, u64= flags); int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value, u64 flags); int bpf_percpu_array_update(struct bpf_map *map, void *key, void *value, @@ -3831,7 +3831,12 @@ bpf_prog_update_insn_ptrs(struct bpf_prog *prog, u32= *offsets, void *image) =20 static inline bool bpf_map_supports_cpu_flags(enum bpf_map_type map_type) { - return false; + switch (map_type) { + case BPF_MAP_TYPE_PERCPU_ARRAY: + return true; + default: + return false; + } } =20 static inline int bpf_map_check_op_flags(struct bpf_map *map, u64 flags, u= 64 allowed_flags) diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index 1eeb31c5b317..67e9e811de3a 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -307,7 +307,7 @@ static void *percpu_array_map_lookup_percpu_elem(struct= bpf_map *map, void *key, return per_cpu_ptr(array->pptrs[index & array->index_mask], cpu); } =20 -int bpf_percpu_array_copy(struct bpf_map *map, void *key, void *value) +int bpf_percpu_array_copy(struct bpf_map *map, void *key, void *value, u64= map_flags) { struct bpf_array *array =3D container_of(map, struct bpf_array, map); u32 index =3D *(u32 *)key; @@ -325,11 +325,18 @@ int bpf_percpu_array_copy(struct bpf_map *map, void *= key, void *value) size =3D array->elem_size; rcu_read_lock(); pptr =3D array->pptrs[index & array->index_mask]; + if (map_flags & BPF_F_CPU) { + cpu =3D map_flags >> 32; + copy_map_value(map, value, per_cpu_ptr(pptr, cpu)); + check_and_init_map_value(map, value); + goto unlock; + } for_each_possible_cpu(cpu) { copy_map_value_long(map, value + off, per_cpu_ptr(pptr, cpu)); check_and_init_map_value(map, value + off); off +=3D size; } +unlock: rcu_read_unlock(); return 0; } @@ -398,10 +405,11 @@ int bpf_percpu_array_update(struct bpf_map *map, void= *key, void *value, struct bpf_array *array =3D container_of(map, struct bpf_array, map); u32 index =3D *(u32 *)key; void __percpu *pptr; - int cpu, off =3D 0; + void *ptr, *val; u32 size; + int cpu; =20 - if (unlikely(map_flags > BPF_EXIST)) + if (unlikely((map_flags & BPF_F_LOCK) || (u32)map_flags > BPF_F_ALL_CPUS)) /* unknown flags */ return -EINVAL; =20 @@ -422,11 +430,20 @@ int bpf_percpu_array_update(struct bpf_map *map, void= *key, void *value, size =3D array->elem_size; rcu_read_lock(); pptr =3D array->pptrs[index & array->index_mask]; + if (map_flags & BPF_F_CPU) { + cpu =3D map_flags >> 32; + ptr =3D per_cpu_ptr(pptr, cpu); + copy_map_value(map, ptr, value); + bpf_obj_free_fields(array->map.record, ptr); + goto unlock; + } for_each_possible_cpu(cpu) { - copy_map_value_long(map, per_cpu_ptr(pptr, cpu), value + off); - bpf_obj_free_fields(array->map.record, per_cpu_ptr(pptr, cpu)); - off +=3D size; + ptr =3D per_cpu_ptr(pptr, cpu); + val =3D (map_flags & BPF_F_ALL_CPUS) ? value : value + size * cpu; + copy_map_value(map, ptr, val); + bpf_obj_free_fields(array->map.record, ptr); } +unlock: rcu_read_unlock(); return 0; } diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 7a1d0ce8fa57..6aa98257fa3f 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -318,7 +318,7 @@ static int bpf_map_copy_value(struct bpf_map *map, void= *key, void *value, map->map_type =3D=3D BPF_MAP_TYPE_LRU_PERCPU_HASH) { err =3D bpf_percpu_hash_copy(map, key, value); } else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_ARRAY) { - err =3D bpf_percpu_array_copy(map, key, value); + err =3D bpf_percpu_array_copy(map, key, value, flags); } else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE) { err =3D bpf_percpu_cgroup_storage_copy(map, key, value); } else if (map->map_type =3D=3D BPF_MAP_TYPE_STACK_TRACE) { --=20 2.51.2 From nobody Mon Dec 1 22:41:53 2025 Received: from out-177.mta0.migadu.com (out-177.mta0.migadu.com [91.218.175.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 7661830595A for ; Wed, 26 Nov 2025 14:52:01 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.177 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168723; cv=none; b=Ns35PE8vJxWJ6H4WvbuWXdYyveHdbscFaC+2CYBwTvp9bqonGxhCE8wFzCKssmXysleg8LiKAH73uqrJ0mXi1JYEkfJqE8uF2X2mpTOWnUHDzzoLiWfz15sMwpJffqgsulHt7CCYCGHYzVOE7gXkqZX0vafQ7DQ4uWWFx99mZDQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168723; c=relaxed/simple; bh=RGRs2ILtPLqcCo7UFzDJ4sYyXt3IuCjZaaU+cqBHZyE=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=oca/b61Lj4JxvEIZvKi17Vi7xfZBlrFZdtXsX2pUJvjQgpjheTYOBrHNC4JS8V0zk8stBQGBXZWo0jUgLZ5Rlhe9blJkT6uyn0GZrmpVIIkgFRA+uLRNHJD+LnCClyQR3NI9t388uVcmcyBRdd04OkdNqlv1L5ybZvBUtkH7S4A= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=JAeXhztk; arc=none smtp.client-ip=91.218.175.177 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="JAeXhztk" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168718; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=gXWUSs+QCIptMMroXrb3t3xD8q7CUAK6Np/i/YAR7wg=; b=JAeXhztkzLBZ8QtYbouJkQgAZrwvJDlX1W1xDSaoQb9SLK5giaxB7GvzTIJoN2D4YWY9Ld HluYxobnsWARvYmCAf81YZ7xh2ygmrsyeJCBVqW71oqHbewP3UFLQL1UxoshHw3iKnjoIW Q8zy+UcbRVGWMCpONo8qnBQgcbo0utY= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 3/7] bpf: Add BPF_F_CPU and BPF_F_ALL_CPUS flags support for percpu_hash and lru_percpu_hash maps Date: Wed, 26 Nov 2025 22:50:35 +0800 Message-ID: <20251126145039.15715-4-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Introduce BPF_F_ALL_CPUS flag support for percpu_hash and lru_percpu_hash maps to allow updating values for all CPUs with a single value for both update_elem and update_batch APIs. Introduce BPF_F_CPU flag support for percpu_hash and lru_percpu_hash maps to allow: * update value for specified CPU for both update_elem and update_batch APIs. * lookup value for specified CPU for both lookup_elem and lookup_batch APIs. The BPF_F_CPU flag is passed via: * map_flags along with embedded cpu info. * elem_flags along with embedded cpu info. Signed-off-by: Leon Hwang --- include/linux/bpf.h | 4 +- kernel/bpf/hashtab.c | 94 ++++++++++++++++++++++++++++++-------------- kernel/bpf/syscall.c | 2 +- 3 files changed, 68 insertions(+), 32 deletions(-) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 01a99e3a3e51..f79d2ae27335 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -2761,7 +2761,7 @@ int map_set_for_each_callback_args(struct bpf_verifie= r_env *env, struct bpf_func_state *caller, struct bpf_func_state *callee); =20 -int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value); +int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value, u64 = flags); int bpf_percpu_array_copy(struct bpf_map *map, void *key, void *value, u64= flags); int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value, u64 flags); @@ -3833,6 +3833,8 @@ static inline bool bpf_map_supports_cpu_flags(enum bp= f_map_type map_type) { switch (map_type) { case BPF_MAP_TYPE_PERCPU_ARRAY: + case BPF_MAP_TYPE_PERCPU_HASH: + case BPF_MAP_TYPE_LRU_PERCPU_HASH: return true; default: return false; diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index c8a9b27f8663..441ff5bc54ac 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -932,7 +932,7 @@ static void free_htab_elem(struct bpf_htab *htab, struc= t htab_elem *l) } =20 static void pcpu_copy_value(struct bpf_htab *htab, void __percpu *pptr, - void *value, bool onallcpus) + void *value, bool onallcpus, u64 map_flags) { void *ptr; =20 @@ -943,19 +943,28 @@ static void pcpu_copy_value(struct bpf_htab *htab, vo= id __percpu *pptr, bpf_obj_free_fields(htab->map.record, ptr); } else { u32 size =3D round_up(htab->map.value_size, 8); - int off =3D 0, cpu; + void *val; + int cpu; + + if (map_flags & BPF_F_CPU) { + cpu =3D map_flags >> 32; + ptr =3D per_cpu_ptr(pptr, cpu); + copy_map_value(&htab->map, ptr, value); + bpf_obj_free_fields(htab->map.record, ptr); + return; + } =20 for_each_possible_cpu(cpu) { ptr =3D per_cpu_ptr(pptr, cpu); - copy_map_value_long(&htab->map, ptr, value + off); + val =3D (map_flags & BPF_F_ALL_CPUS) ? value : value + size * cpu; + copy_map_value(&htab->map, ptr, val); bpf_obj_free_fields(htab->map.record, ptr); - off +=3D size; } } } =20 static void pcpu_init_value(struct bpf_htab *htab, void __percpu *pptr, - void *value, bool onallcpus) + void *value, bool onallcpus, u64 map_flags) { /* When not setting the initial value on all cpus, zero-fill element * values for other cpus. Otherwise, bpf program has no way to ensure @@ -973,7 +982,7 @@ static void pcpu_init_value(struct bpf_htab *htab, void= __percpu *pptr, zero_map_value(&htab->map, per_cpu_ptr(pptr, cpu)); } } else { - pcpu_copy_value(htab, pptr, value, onallcpus); + pcpu_copy_value(htab, pptr, value, onallcpus, map_flags); } } =20 @@ -985,7 +994,7 @@ static bool fd_htab_map_needs_adjust(const struct bpf_h= tab *htab) static struct htab_elem *alloc_htab_elem(struct bpf_htab *htab, void *key, void *value, u32 key_size, u32 hash, bool percpu, bool onallcpus, - struct htab_elem *old_elem) + struct htab_elem *old_elem, u64 map_flags) { u32 size =3D htab->map.value_size; bool prealloc =3D htab_is_prealloc(htab); @@ -1043,7 +1052,7 @@ static struct htab_elem *alloc_htab_elem(struct bpf_h= tab *htab, void *key, pptr =3D *(void __percpu **)ptr; } =20 - pcpu_init_value(htab, pptr, value, onallcpus); + pcpu_init_value(htab, pptr, value, onallcpus, map_flags); =20 if (!prealloc) htab_elem_set_ptr(l_new, key_size, pptr); @@ -1147,7 +1156,7 @@ static long htab_map_update_elem(struct bpf_map *map,= void *key, void *value, } =20 l_new =3D alloc_htab_elem(htab, key, value, key_size, hash, false, false, - l_old); + l_old, map_flags); if (IS_ERR(l_new)) { /* all pre-allocated elements are in use or memory exhausted */ ret =3D PTR_ERR(l_new); @@ -1249,6 +1258,15 @@ static long htab_lru_map_update_elem(struct bpf_map = *map, void *key, void *value return ret; } =20 +static int htab_map_check_update_flags(bool onallcpus, u64 map_flags) +{ + if (unlikely(!onallcpus && map_flags > BPF_EXIST)) + return -EINVAL; + if (unlikely(onallcpus && ((map_flags & BPF_F_LOCK) || (u32)map_flags > B= PF_F_ALL_CPUS))) + return -EINVAL; + return 0; +} + static long htab_map_update_elem_in_place(struct bpf_map *map, void *key, void *value, u64 map_flags, bool percpu, bool onallcpus) @@ -1262,9 +1280,9 @@ static long htab_map_update_elem_in_place(struct bpf_= map *map, void *key, u32 key_size, hash; int ret; =20 - if (unlikely(map_flags > BPF_EXIST)) - /* unknown flags */ - return -EINVAL; + ret =3D htab_map_check_update_flags(onallcpus, map_flags); + if (unlikely(ret)) + return ret; =20 WARN_ON_ONCE(!bpf_rcu_lock_held()); =20 @@ -1289,7 +1307,7 @@ static long htab_map_update_elem_in_place(struct bpf_= map *map, void *key, /* Update value in-place */ if (percpu) { pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size), - value, onallcpus); + value, onallcpus, map_flags); } else { void **inner_map_pptr =3D htab_elem_value(l_old, key_size); =20 @@ -1298,7 +1316,7 @@ static long htab_map_update_elem_in_place(struct bpf_= map *map, void *key, } } else { l_new =3D alloc_htab_elem(htab, key, value, key_size, - hash, percpu, onallcpus, NULL); + hash, percpu, onallcpus, NULL, map_flags); if (IS_ERR(l_new)) { ret =3D PTR_ERR(l_new); goto err; @@ -1324,9 +1342,9 @@ static long __htab_lru_percpu_map_update_elem(struct = bpf_map *map, void *key, u32 key_size, hash; int ret; =20 - if (unlikely(map_flags > BPF_EXIST)) - /* unknown flags */ - return -EINVAL; + ret =3D htab_map_check_update_flags(onallcpus, map_flags); + if (unlikely(ret)) + return ret; =20 WARN_ON_ONCE(!bpf_rcu_lock_held()); =20 @@ -1363,10 +1381,10 @@ static long __htab_lru_percpu_map_update_elem(struc= t bpf_map *map, void *key, =20 /* per-cpu hash map can update value in-place */ pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size), - value, onallcpus); + value, onallcpus, map_flags); } else { pcpu_init_value(htab, htab_elem_get_ptr(l_new, key_size), - value, onallcpus); + value, onallcpus, map_flags); hlist_nulls_add_head_rcu(&l_new->hash_node, head); l_new =3D NULL; } @@ -1678,9 +1696,9 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *ma= p, void __user *ukeys =3D u64_to_user_ptr(attr->batch.keys); void __user *ubatch =3D u64_to_user_ptr(attr->batch.in_batch); u32 batch, max_count, size, bucket_size, map_id; + u64 elem_map_flags, map_flags, allowed_flags; u32 bucket_cnt, total, key_size, value_size; struct htab_elem *node_to_free =3D NULL; - u64 elem_map_flags, map_flags; struct hlist_nulls_head *head; struct hlist_nulls_node *n; unsigned long flags =3D 0; @@ -1690,9 +1708,12 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *m= ap, int ret =3D 0; =20 elem_map_flags =3D attr->batch.elem_flags; - if ((elem_map_flags & ~BPF_F_LOCK) || - ((elem_map_flags & BPF_F_LOCK) && !btf_record_has_field(map->record, = BPF_SPIN_LOCK))) - return -EINVAL; + allowed_flags =3D BPF_F_LOCK; + if (!do_delete && is_percpu) + allowed_flags |=3D BPF_F_CPU; + ret =3D bpf_map_check_op_flags(map, elem_map_flags, allowed_flags); + if (ret) + return ret; =20 map_flags =3D attr->batch.flags; if (map_flags) @@ -1715,7 +1736,7 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *ma= p, key_size =3D htab->map.key_size; value_size =3D htab->map.value_size; size =3D round_up(value_size, 8); - if (is_percpu) + if (is_percpu && !(elem_map_flags & BPF_F_CPU)) value_size =3D size * num_possible_cpus(); total =3D 0; /* while experimenting with hash tables with sizes ranging from 10 to @@ -1798,10 +1819,17 @@ __htab_map_lookup_and_delete_batch(struct bpf_map *= map, void __percpu *pptr; =20 pptr =3D htab_elem_get_ptr(l, map->key_size); - for_each_possible_cpu(cpu) { - copy_map_value_long(&htab->map, dst_val + off, per_cpu_ptr(pptr, cpu)); - check_and_init_map_value(&htab->map, dst_val + off); - off +=3D size; + if (elem_map_flags & BPF_F_CPU) { + cpu =3D elem_map_flags >> 32; + copy_map_value(&htab->map, dst_val, per_cpu_ptr(pptr, cpu)); + check_and_init_map_value(&htab->map, dst_val); + } else { + for_each_possible_cpu(cpu) { + copy_map_value_long(&htab->map, dst_val + off, + per_cpu_ptr(pptr, cpu)); + check_and_init_map_value(&htab->map, dst_val + off); + off +=3D size; + } } } else { value =3D htab_elem_value(l, key_size); @@ -2357,7 +2385,7 @@ static void *htab_lru_percpu_map_lookup_percpu_elem(s= truct bpf_map *map, void *k return NULL; } =20 -int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value) +int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value, u64 = map_flags) { struct htab_elem *l; void __percpu *pptr; @@ -2374,16 +2402,22 @@ int bpf_percpu_hash_copy(struct bpf_map *map, void = *key, void *value) l =3D __htab_map_lookup_elem(map, key); if (!l) goto out; + ret =3D 0; /* We do not mark LRU map element here in order to not mess up * eviction heuristics when user space does a map walk. */ pptr =3D htab_elem_get_ptr(l, map->key_size); + if (map_flags & BPF_F_CPU) { + cpu =3D map_flags >> 32; + copy_map_value(map, value, per_cpu_ptr(pptr, cpu)); + check_and_init_map_value(map, value); + goto out; + } for_each_possible_cpu(cpu) { copy_map_value_long(map, value + off, per_cpu_ptr(pptr, cpu)); check_and_init_map_value(map, value + off); off +=3D size; } - ret =3D 0; out: rcu_read_unlock(); return ret; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 6aa98257fa3f..94ac63b1115f 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -316,7 +316,7 @@ static int bpf_map_copy_value(struct bpf_map *map, void= *key, void *value, bpf_disable_instrumentation(); if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_HASH || map->map_type =3D=3D BPF_MAP_TYPE_LRU_PERCPU_HASH) { - err =3D bpf_percpu_hash_copy(map, key, value); + err =3D bpf_percpu_hash_copy(map, key, value, flags); } else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_ARRAY) { err =3D bpf_percpu_array_copy(map, key, value, flags); } else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE) { --=20 2.51.2 From nobody Mon Dec 1 22:41:53 2025 Received: from out-186.mta0.migadu.com (out-186.mta0.migadu.com [91.218.175.186]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C59BB305048 for ; Wed, 26 Nov 2025 14:52:12 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.186 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168734; cv=none; b=b/nmOf2mW8u0DALMXtnxesxWxrACilegaq4bOoqaOgRt+CXRJZrSUZxPwSsXU6Z/0hqF1ApKc2XKJwgDGgLBfnM3MBRNOegplbMTjWzY3RC6VxsHDIGiCpk5/SyEE9TpIsa0hwDQ9zYEIGnrZfrOkMVWh/geq6Ul+SkTFwaHG3o= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168734; c=relaxed/simple; bh=SZ9BjoQFDYKINn07VcspgXWVdM9dxtKA1Gv6JFK8Y5o=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=E8FoBB3rJR16PkLhcGR41rOjoUtdVkJR3StUjotkNDhWHsOiKJuyChRifdnnYEJzUeeSs8BG13+eDHfHhKFdOa7uwH0X+l6gqAHjh73UIFmamQs/T30FgK4OH/YypE06/4W8ouuoThCKxqyVrAoQmsi3WkJKdUyFjJjcVKD4jAQ= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=GIU7cTF6; arc=none smtp.client-ip=91.218.175.186 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="GIU7cTF6" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168729; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=pTgwqBgBsAEUNrMYXmgH0c/Rvt/LDiv6qDBO69l3Lvo=; b=GIU7cTF6BEIB19xr9G2VgzlrID2kRyYTLeLpshQ+NZDITLBgss5B66VA2iiU8zJ9KFlJxh KfrS766zVMHCxvb8AmtA8O0neNT7Us0BrCuGfUsfogJ/xjzrAhtHMPYIJ9u/wwvG5MJOrs YXJv9E6gfFOGiqhqsZbVmYl3LnchbaQ= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 4/7] bpf: Copy map value using copy_map_value_long for percpu_cgroup_storage maps Date: Wed, 26 Nov 2025 22:50:36 +0800 Message-ID: <20251126145039.15715-5-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Copy map value using 'copy_map_value_long()'. It's to keep consistent style with the way of other percpu maps. No functional change intended. Signed-off-by: Leon Hwang --- kernel/bpf/local_storage.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c index c93a756e035c..2ab4b60ffe61 100644 --- a/kernel/bpf/local_storage.c +++ b/kernel/bpf/local_storage.c @@ -200,8 +200,7 @@ int bpf_percpu_cgroup_storage_copy(struct bpf_map *_map= , void *key, */ size =3D round_up(_map->value_size, 8); for_each_possible_cpu(cpu) { - bpf_long_memcpy(value + off, - per_cpu_ptr(storage->percpu_buf, cpu), size); + copy_map_value_long(_map, value + off, per_cpu_ptr(storage->percpu_buf, = cpu)); off +=3D size; } rcu_read_unlock(); @@ -234,8 +233,7 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *_m= ap, void *key, */ size =3D round_up(_map->value_size, 8); for_each_possible_cpu(cpu) { - bpf_long_memcpy(per_cpu_ptr(storage->percpu_buf, cpu), - value + off, size); + copy_map_value_long(_map, per_cpu_ptr(storage->percpu_buf, cpu), value += off); off +=3D size; } rcu_read_unlock(); --=20 2.51.2 From nobody Mon Dec 1 22:41:53 2025 Received: from out-188.mta0.migadu.com (out-188.mta0.migadu.com [91.218.175.188]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D0392303CA4 for ; Wed, 26 Nov 2025 14:52:22 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.188 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168744; cv=none; b=c/RnvqdGXCaloZz5yAZvISS8rIPEXCNLETMj3GIjlcP9XKct6HTVNpg+c48Ge+6LCb3/W33GvC6QIaVEvD1eUKAEkbVePnoQzrBwnfoP0Th95T7P28edkm++7LVJC50CewXgVyq+guvwv0mIakZ9f9sHk8Xy2PAkmukph+qO4RU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168744; c=relaxed/simple; bh=LlYYr6YXrR6QEPCnnQAYyWnfYGASwCRYkM8mkTO03Pw=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=CAVtf3HAiFJMEmIAtlyYB/Zm4vo8q8eiy+tji+P1nl8OfBmOBPL2+Ywzw7AEtp9qVl2lHUUHHIvPl/pZzdM6EgCtp0XHjs/kZ6fdpDOsGbu2A2Mg2ejAFMaiYPDliRdZ3P0F8H/dPMvo1uODo02yX0xKUw3P8O5cStdTojQMmq4= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=Gqp8BOey; arc=none smtp.client-ip=91.218.175.188 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="Gqp8BOey" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168740; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=X2X+KLzyvJwhZKKjugRp6VQbzHu5EbCujgdbtaP+foU=; b=Gqp8BOeyOmHFE2sY2163AstDWBoBdYtg7wJBoMjHPAuju8qBi3Ggej9OzDuM0+XEdoh+TH bR+Kwr+pV9jTZQKH6+ZDEsPUbKSPmL0cd3rjWVRrdDxmwCTZAdbMTAZSFeMKEyEYSMA7BY sz2GVhI1GmzjP2cxvGgyJwNFhsSFfW8= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 5/7] bpf: Add BPF_F_CPU and BPF_F_ALL_CPUS flags support for percpu_cgroup_storage maps Date: Wed, 26 Nov 2025 22:50:37 +0800 Message-ID: <20251126145039.15715-6-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Introduce BPF_F_ALL_CPUS flag support for percpu_cgroup_storage maps to allow updating values for all CPUs with a single value for update_elem API. Introduce BPF_F_CPU flag support for percpu_cgroup_storage maps to allow: * update value for specified CPU for update_elem API. * lookup value for specified CPU for lookup_elem API. The BPF_F_CPU flag is passed via map_flags along with embedded cpu info. Signed-off-by: Leon Hwang --- include/linux/bpf-cgroup.h | 4 ++-- include/linux/bpf.h | 1 + kernel/bpf/local_storage.c | 23 ++++++++++++++++++----- kernel/bpf/syscall.c | 2 +- 4 files changed, 22 insertions(+), 8 deletions(-) diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index aedf573bdb42..013f4db9903f 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -172,7 +172,7 @@ void bpf_cgroup_storage_link(struct bpf_cgroup_storage = *storage, void bpf_cgroup_storage_unlink(struct bpf_cgroup_storage *storage); int bpf_cgroup_storage_assign(struct bpf_prog_aux *aux, struct bpf_map *ma= p); =20 -int bpf_percpu_cgroup_storage_copy(struct bpf_map *map, void *key, void *v= alue); +int bpf_percpu_cgroup_storage_copy(struct bpf_map *map, void *key, void *v= alue, u64 flags); int bpf_percpu_cgroup_storage_update(struct bpf_map *map, void *key, void *value, u64 flags); =20 @@ -467,7 +467,7 @@ static inline struct bpf_cgroup_storage *bpf_cgroup_sto= rage_alloc( static inline void bpf_cgroup_storage_free( struct bpf_cgroup_storage *storage) {} static inline int bpf_percpu_cgroup_storage_copy(struct bpf_map *map, void= *key, - void *value) { + void *value, u64 flags) { return 0; } static inline int bpf_percpu_cgroup_storage_update(struct bpf_map *map, diff --git a/include/linux/bpf.h b/include/linux/bpf.h index f79d2ae27335..9e756db5e132 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -3835,6 +3835,7 @@ static inline bool bpf_map_supports_cpu_flags(enum bp= f_map_type map_type) case BPF_MAP_TYPE_PERCPU_ARRAY: case BPF_MAP_TYPE_PERCPU_HASH: case BPF_MAP_TYPE_LRU_PERCPU_HASH: + case BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE: return true; default: return false; diff --git a/kernel/bpf/local_storage.c b/kernel/bpf/local_storage.c index 2ab4b60ffe61..1ccbf28b2ad9 100644 --- a/kernel/bpf/local_storage.c +++ b/kernel/bpf/local_storage.c @@ -180,7 +180,7 @@ static long cgroup_storage_update_elem(struct bpf_map *= map, void *key, } =20 int bpf_percpu_cgroup_storage_copy(struct bpf_map *_map, void *key, - void *value) + void *value, u64 map_flags) { struct bpf_cgroup_storage_map *map =3D map_to_storage(_map); struct bpf_cgroup_storage *storage; @@ -198,11 +198,17 @@ int bpf_percpu_cgroup_storage_copy(struct bpf_map *_m= ap, void *key, * access 'value_size' of them, so copying rounded areas * will not leak any kernel data */ + if (map_flags & BPF_F_CPU) { + cpu =3D map_flags >> 32; + copy_map_value(_map, value, per_cpu_ptr(storage->percpu_buf, cpu)); + goto unlock; + } size =3D round_up(_map->value_size, 8); for_each_possible_cpu(cpu) { copy_map_value_long(_map, value + off, per_cpu_ptr(storage->percpu_buf, = cpu)); off +=3D size; } +unlock: rcu_read_unlock(); return 0; } @@ -212,10 +218,11 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *= _map, void *key, { struct bpf_cgroup_storage_map *map =3D map_to_storage(_map); struct bpf_cgroup_storage *storage; - int cpu, off =3D 0; + void *val; u32 size; + int cpu; =20 - if (map_flags !=3D BPF_ANY && map_flags !=3D BPF_EXIST) + if ((u32)map_flags & ~(BPF_ANY | BPF_EXIST | BPF_F_CPU | BPF_F_ALL_CPUS)) return -EINVAL; =20 rcu_read_lock(); @@ -231,11 +238,17 @@ int bpf_percpu_cgroup_storage_update(struct bpf_map *= _map, void *key, * returned or zeros which were zero-filled by percpu_alloc, * so no kernel data leaks possible */ + if (map_flags & BPF_F_CPU) { + cpu =3D map_flags >> 32; + copy_map_value(_map, per_cpu_ptr(storage->percpu_buf, cpu), value); + goto unlock; + } size =3D round_up(_map->value_size, 8); for_each_possible_cpu(cpu) { - copy_map_value_long(_map, per_cpu_ptr(storage->percpu_buf, cpu), value += off); - off +=3D size; + val =3D (map_flags & BPF_F_ALL_CPUS) ? value : value + size * cpu; + copy_map_value(_map, per_cpu_ptr(storage->percpu_buf, cpu), val); } +unlock: rcu_read_unlock(); return 0; } diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 94ac63b1115f..d8c2336124b6 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -320,7 +320,7 @@ static int bpf_map_copy_value(struct bpf_map *map, void= *key, void *value, } else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_ARRAY) { err =3D bpf_percpu_array_copy(map, key, value, flags); } else if (map->map_type =3D=3D BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE) { - err =3D bpf_percpu_cgroup_storage_copy(map, key, value); + err =3D bpf_percpu_cgroup_storage_copy(map, key, value, flags); } else if (map->map_type =3D=3D BPF_MAP_TYPE_STACK_TRACE) { err =3D bpf_stackmap_extract(map, key, value, false); } else if (IS_FD_ARRAY(map) || IS_FD_PROG_ARRAY(map)) { --=20 2.51.2 From nobody Mon Dec 1 22:41:53 2025 Received: from out-170.mta0.migadu.com (out-170.mta0.migadu.com [91.218.175.170]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 12C88305068 for ; Wed, 26 Nov 2025 14:52:32 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.170 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168753; cv=none; b=pi9OMJEFG2p6Fd2jkFRPRyTl3JR3rrnBb4prLAs7xZu55sAZCiEkJqynOjYpDGkQUmL4LmIk/8nultP+VP/h+PULu+7/VrXEWcxoC5QCoEp8+zBW/pK7T+qRWGNGiBYPhCpyY+HYSLsnCH3serIcM+GcBSYlsR2T9poZtbiO1Zo= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168753; c=relaxed/simple; bh=tv6u8+8HiGWGT3wL1c8GXPJ+6n3YuwfoeZWo5+YnnG0=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=Aih8EupN4KUyULK1dPcuLSIFGjffn5uUIy1T7thYrW6QMx1Y1qOr1r+Uya+hoidifqPx07f6SYkRa1blrKZfQOZHLKWyOdCESDM6BCMMWgx99n5bxSyF9gvANWwCvxrEf14HCV7fgMSDXEGDjygdH0sJIb3bz0UdyJFXEqfE02Y= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=I37A+rCL; arc=none smtp.client-ip=91.218.175.170 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="I37A+rCL" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168750; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=eyVpsABIiydOnG1kb3o6PrVSMcQesGuFwWpVoehPpLc=; b=I37A+rCLQZBOVUKmlVEmGv71sQCyXJLc0fY2bjGzA0PKa312CVorS0uOTFjDtdWeMVtJi8 AhzszCYoAdm8X1/qh7cr5xpzTJYFk3xdvCm3n0NRQ1jTGmXUjOZeeQIsDUSLXIYVAxy8rC GQNTzoPp18j9HO9Uv7fLglVwtEIKlO0= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 6/7] libbpf: Add BPF_F_CPU and BPF_F_ALL_CPUS flags support for percpu maps Date: Wed, 26 Nov 2025 22:50:38 +0800 Message-ID: <20251126145039.15715-7-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" 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. Acked-by: Andrii Nakryiko Signed-off-by: Leon Hwang --- tools/lib/bpf/bpf.h | 8 ++++++++ tools/lib/bpf/libbpf.c | 26 ++++++++++++++++++++------ tools/lib/bpf/libbpf.h | 21 ++++++++------------- 3 files changed, 36 insertions(+), 19 deletions(-) diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h index e983a3e40d61..ffd93feffd71 100644 --- a/tools/lib/bpf/bpf.h +++ b/tools/lib/bpf/bpf.h @@ -289,6 +289,14 @@ LIBBPF_API int bpf_map_lookup_and_delete_batch(int fd,= void *in_batch, * Update spin_lock-ed map elements. This must be * specified if the map value contains a spinlock. * + * **BPF_F_CPU** + * As for percpu maps, update value on the specified CPU. And the cpu + * info is embedded into the high 32 bits of **opts->elem_flags**. + * + * **BPF_F_ALL_CPUS** + * As for percpu maps, update value across all CPUs. This flag cannot + * be used with BPF_F_CPU at the same time. + * * @param fd BPF map file descriptor * @param keys pointer to an array of *count* keys * @param values pointer to an array of *count* values diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 706e7481bdf6..65b9b5e95544 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -10913,7 +10913,7 @@ bpf_object__find_map_fd_by_name(const struct bpf_ob= ject *obj, const char *name) } =20 static int validate_map_op(const struct bpf_map *map, size_t key_sz, - size_t value_sz, bool check_value_sz) + size_t value_sz, bool check_value_sz, __u64 flags) { if (!map_is_created(map)) /* map is not yet created */ return -ENOENT; @@ -10940,6 +10940,20 @@ static int validate_map_op(const struct bpf_map *m= ap, size_t key_sz, int num_cpu =3D libbpf_num_possible_cpus(); size_t elem_sz =3D roundup(map->def.value_size, 8); =20 + if (flags & (BPF_F_CPU | BPF_F_ALL_CPUS)) { + if ((flags & BPF_F_CPU) && (flags & BPF_F_ALL_CPUS)) { + pr_warn("map '%s': BPF_F_CPU and BPF_F_ALL_CPUS are mutually exclusive= \n", + map->name); + return -EINVAL; + } + if (map->def.value_size !=3D value_sz) { + pr_warn("map '%s': unexpected value size %zu provided for either BPF_F= _CPU or BPF_F_ALL_CPUS, expected %u\n", + map->name, value_sz, map->def.value_size); + return -EINVAL; + } + break; + } + if (value_sz !=3D num_cpu * elem_sz) { pr_warn("map '%s': unexpected value size %zu provided for per-CPU map, = expected %d * %zu =3D %zd\n", map->name, value_sz, num_cpu, elem_sz, num_cpu * elem_sz); @@ -10964,7 +10978,7 @@ int bpf_map__lookup_elem(const struct bpf_map *map, { int err; =20 - err =3D validate_map_op(map, key_sz, value_sz, true); + err =3D validate_map_op(map, key_sz, value_sz, true, flags); if (err) return libbpf_err(err); =20 @@ -10977,7 +10991,7 @@ int bpf_map__update_elem(const struct bpf_map *map, { int err; =20 - err =3D validate_map_op(map, key_sz, value_sz, true); + err =3D validate_map_op(map, key_sz, value_sz, true, flags); if (err) return libbpf_err(err); =20 @@ -10989,7 +11003,7 @@ int bpf_map__delete_elem(const struct bpf_map *map, { int err; =20 - err =3D validate_map_op(map, key_sz, 0, false /* check_value_sz */); + err =3D validate_map_op(map, key_sz, 0, false /* check_value_sz */, flags= ); if (err) return libbpf_err(err); =20 @@ -11002,7 +11016,7 @@ int bpf_map__lookup_and_delete_elem(const struct bp= f_map *map, { int err; =20 - err =3D validate_map_op(map, key_sz, value_sz, true); + err =3D validate_map_op(map, key_sz, value_sz, true, flags); if (err) return libbpf_err(err); =20 @@ -11014,7 +11028,7 @@ int bpf_map__get_next_key(const struct bpf_map *map, { int err; =20 - err =3D validate_map_op(map, key_sz, 0, false /* check_value_sz */); + err =3D validate_map_op(map, key_sz, 0, false /* check_value_sz */, 0); if (err) return libbpf_err(err); =20 diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h index 65e68e964b89..414e8a873380 100644 --- a/tools/lib/bpf/libbpf.h +++ b/tools/lib/bpf/libbpf.h @@ -1200,12 +1200,13 @@ LIBBPF_API struct bpf_map *bpf_map__inner_map(struc= t bpf_map *map); * @param key_sz size in bytes of key data, needs to match BPF map definit= ion's **key_size** * @param value pointer to memory in which looked up value will be stored * @param value_sz size in byte of value data memory; it has to match BPF = map - * definition's **value_size**. For per-CPU BPF maps value size has to be - * a product of BPF map value size and number of possible CPUs in the syst= em - * (could be fetched with **libbpf_num_possible_cpus()**). Note also that = for - * per-CPU values value size has to be aligned up to closest 8 bytes for - * alignment reasons, so expected size is: `round_up(value_size, 8) - * * libbpf_num_possible_cpus()`. + * definition's **value_size**. For per-CPU BPF maps, value size can be + * `value_size` if either **BPF_F_CPU** or **BPF_F_ALL_CPUS** is specified + * in **flags**, otherwise a product of BPF map value size and number of + * possible CPUs in the system (could be fetched with + * **libbpf_num_possible_cpus()**). Note also that for per-CPU values value + * size has to be aligned up to closest 8 bytes, so expected size is: + * `round_up(value_size, 8) * libbpf_num_possible_cpus()`. * @param flags extra flags passed to kernel for this operation * @return 0, on success; negative error, otherwise * @@ -1223,13 +1224,7 @@ LIBBPF_API int bpf_map__lookup_elem(const struct bpf= _map *map, * @param key pointer to memory containing bytes of the key * @param key_sz size in bytes of key data, needs to match BPF map definit= ion's **key_size** * @param value pointer to memory containing bytes of the value - * @param value_sz size in byte of value data memory; it has to match BPF = map - * definition's **value_size**. For per-CPU BPF maps value size has to be - * a product of BPF map value size and number of possible CPUs in the syst= em - * (could be fetched with **libbpf_num_possible_cpus()**). Note also that = for - * per-CPU values value size has to be aligned up to closest 8 bytes for - * alignment reasons, so expected size is: `round_up(value_size, 8) - * * libbpf_num_possible_cpus()`. + * @param value_sz refer to **bpf_map__lookup_elem**'s description.' * @param flags extra flags passed to kernel for this operation * @return 0, on success; negative error, otherwise * --=20 2.51.2 From nobody Mon Dec 1 22:41:53 2025 Received: from out-174.mta0.migadu.com (out-174.mta0.migadu.com [91.218.175.174]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id F0636308F1D for ; Wed, 26 Nov 2025 14:52:42 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=91.218.175.174 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168765; cv=none; b=WQY5PxaOgUkzmDh+n83vzqAedZwERm1gkrZSe9KnlLkBRIWorNPPy3ekJSb2s9XBUy42+l6XRh7td0rIO7FZyAiJRVmNY32vqhZFckaJ94mp1JAHwuER7o/rmD5YDU8LBatP750NMxmmHTAuqjzDIJlKCsbUbKQXRMVm94OYRsI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1764168765; c=relaxed/simple; bh=ObBVKeOuW/rZyhKKNHsmLaLzvEHYqS7q/OKIbcl6Ixs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=hX2ZobPq00CI7kXKCkIUpm9m9biUBsOOp29cTfJqApOieJcJT54PBlbbKWaj9RfOegwwdrboS9qywrlSAoxVUteId0dWW53NIdtl+14+DUSXcqCYXoUlzLgb+YZYhf5L+dXrIHLtIe6wT/XPx/q/7CxeRXBCgl/VFiMa8SPImV0= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=FFZnVk4t; arc=none smtp.client-ip=91.218.175.174 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="FFZnVk4t" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1764168760; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=zQW7AL/CsZzRnWa2E56YrYrrhZjrdykemufrqeiU4zk=; b=FFZnVk4tmc5rw/FCp1XYuYOeZpmGBgX+B10LW9Knn8u1cyeg8Kqfjq173LGa1r7LTyII+5 TC5hoMkG1602tgSs0W/wDvXDybq0PjhX6OWEXlLx3Tn0NtlW8TocO5BoiXNTdCOyZ/m+qo ayPnsUmieNXQkHJkWHtHJjefHCYnKbo= From: Leon Hwang To: bpf@vger.kernel.org Cc: ast@kernel.org, andrii@kernel.org, daniel@iogearbox.net, jolsa@kernel.org, yonghong.song@linux.dev, song@kernel.org, eddyz87@gmail.com, dxu@dxuuu.xyz, deso@posteo.net, martin.lau@linux.dev, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@fomichev.me, haoluo@google.com, shuah@kernel.org, kerneljasonxing@gmail.com, chen.dylane@linux.dev, willemb@google.com, paul.chaignon@gmail.com, a.s.protopopov@gmail.com, memxor@gmail.com, yatsenko@meta.com, tklauser@distanz.ch, leon.hwang@linux.dev, kernel-patches-bot@fb.com, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v12 7/7] selftests/bpf: Add cases to test BPF_F_CPU and BPF_F_ALL_CPUS flags Date: Wed, 26 Nov 2025 22:50:39 +0800 Message-ID: <20251126145039.15715-8-leon.hwang@linux.dev> In-Reply-To: <20251126145039.15715-1-leon.hwang@linux.dev> References: <20251126145039.15715-1-leon.hwang@linux.dev> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Add test coverage for the new BPF_F_CPU and BPF_F_ALL_CPUS flags support in percpu maps. The following APIs are exercised: * bpf_map_update_batch() * bpf_map_lookup_batch() * bpf_map_update_elem() * bpf_map__update_elem() * bpf_map_lookup_elem_flags() * bpf_map__lookup_elem() For lru_percpu_hash map, set max_entries to 'libbpf_num_possible_cpus() + 1' and only use the first 'libbpf_num_possible_cpus()' entries. This ensures a spare entry is always available in the LRU free list, avoiding eviction. When updating an existing key in lru_percpu_hash map: 1. l_new =3D prealloc_lru_pop(); /* Borrow from free list */ 2. l_old =3D lookup_elem_raw(); /* Found, key exists */ 3. pcpu_copy_value(); /* In-place update */ 4. bpf_lru_push_free(); /* Return l_new to free list */ Also add negative tests to verify that non-percpu array and hash maps reject the BPF_F_CPU and BPF_F_ALL_CPUS flags. Signed-off-by: Leon Hwang --- .../selftests/bpf/prog_tests/percpu_alloc.c | 328 ++++++++++++++++++ .../selftests/bpf/progs/percpu_alloc_array.c | 32 ++ 2 files changed, 360 insertions(+) diff --git a/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c b/tools/= testing/selftests/bpf/prog_tests/percpu_alloc.c index 343da65864d6..c1d0949f093f 100644 --- a/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c +++ b/tools/testing/selftests/bpf/prog_tests/percpu_alloc.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #include +#include "cgroup_helpers.h" #include "percpu_alloc_array.skel.h" #include "percpu_alloc_cgrp_local_storage.skel.h" #include "percpu_alloc_fail.skel.h" @@ -115,6 +116,321 @@ static void test_failure(void) { RUN_TESTS(percpu_alloc_fail); } =20 +static void test_percpu_map_op_cpu_flag(struct bpf_map *map, void *keys, s= ize_t key_sz, u32 entries, + int nr_cpus, bool test_batch) +{ + size_t value_sz =3D sizeof(u32), value_sz_cpus, value_sz_total; + u32 *values =3D NULL, *values_percpu =3D NULL; + const u32 value =3D 0xDEADC0DE; + int i, j, cpu, map_fd, err; + u64 batch =3D 0, flags; + void *values_row; + u32 count, v; + LIBBPF_OPTS(bpf_map_batch_opts, batch_opts); + + value_sz_cpus =3D value_sz * nr_cpus; + values =3D calloc(entries, value_sz_cpus); + if (!ASSERT_OK_PTR(values, "calloc values")) + return; + + values_percpu =3D calloc(entries, roundup(value_sz, 8) * nr_cpus); + if (!ASSERT_OK_PTR(values_percpu, "calloc values_percpu")) { + free(values); + return; + } + + value_sz_total =3D value_sz_cpus * entries; + memset(values, 0, value_sz_total); + + map_fd =3D bpf_map__fd(map); + flags =3D BPF_F_CPU | BPF_F_ALL_CPUS; + err =3D bpf_map_lookup_elem_flags(map_fd, keys, values, flags); + if (!ASSERT_ERR(err, "bpf_map_lookup_elem_flags cpu|all_cpus")) + goto out; + + err =3D bpf_map_update_elem(map_fd, keys, values, flags); + if (!ASSERT_ERR(err, "bpf_map_update_elem cpu|all_cpus")) + goto out; + + flags =3D BPF_F_ALL_CPUS; + err =3D bpf_map_lookup_elem_flags(map_fd, keys, values, flags); + if (!ASSERT_ERR(err, "bpf_map_lookup_elem_flags all_cpus")) + goto out; + + flags =3D BPF_F_LOCK | BPF_F_CPU; + err =3D bpf_map_lookup_elem_flags(map_fd, keys, values, flags); + if (!ASSERT_ERR(err, "bpf_map_lookup_elem_flags BPF_F_LOCK")) + goto out; + + flags =3D BPF_F_LOCK | BPF_F_ALL_CPUS; + err =3D bpf_map_update_elem(map_fd, keys, values, flags); + if (!ASSERT_ERR(err, "bpf_map_update_elem BPF_F_LOCK")) + goto out; + + flags =3D (u64)nr_cpus << 32 | BPF_F_CPU; + err =3D bpf_map_update_elem(map_fd, keys, values, flags); + if (!ASSERT_EQ(err, -ERANGE, "bpf_map_update_elem -ERANGE")) + goto out; + + err =3D bpf_map__update_elem(map, keys, key_sz, values, value_sz, flags); + if (!ASSERT_EQ(err, -ERANGE, "bpf_map__update_elem -ERANGE")) + goto out; + + err =3D bpf_map_lookup_elem_flags(map_fd, keys, values, flags); + if (!ASSERT_EQ(err, -ERANGE, "bpf_map_lookup_elem_flags -ERANGE")) + goto out; + + err =3D bpf_map__lookup_elem(map, keys, key_sz, values, value_sz, flags); + if (!ASSERT_EQ(err, -ERANGE, "bpf_map__lookup_elem -ERANGE")) + goto out; + + for (cpu =3D 0; cpu < nr_cpus; cpu++) { + /* clear value on all cpus */ + values[0] =3D 0; + flags =3D BPF_F_ALL_CPUS; + for (i =3D 0; i < entries; i++) { + err =3D bpf_map__update_elem(map, keys + i * key_sz, key_sz, values, + value_sz, flags); + if (!ASSERT_OK(err, "bpf_map__update_elem all_cpus")) + goto out; + } + + /* update value on specified cpu */ + for (i =3D 0; i < entries; i++) { + values[0] =3D value; + flags =3D (u64)cpu << 32 | BPF_F_CPU; + err =3D bpf_map__update_elem(map, keys + i * key_sz, key_sz, values, + value_sz, flags); + if (!ASSERT_OK(err, "bpf_map__update_elem specified cpu")) + goto out; + + /* lookup then check value on CPUs */ + for (j =3D 0; j < nr_cpus; j++) { + flags =3D (u64)j << 32 | BPF_F_CPU; + err =3D bpf_map__lookup_elem(map, keys + i * key_sz, key_sz, values, + value_sz, flags); + if (!ASSERT_OK(err, "bpf_map__lookup_elem specified cpu")) + goto out; + if (!ASSERT_EQ(values[0], j !=3D cpu ? 0 : value, + "bpf_map__lookup_elem value on specified cpu")) + goto out; + } + } + } + + if (!test_batch) + goto out; + + count =3D entries; + batch_opts.elem_flags =3D (u64)nr_cpus << 32 | BPF_F_CPU; + err =3D bpf_map_update_batch(map_fd, keys, values, &count, &batch_opts); + if (!ASSERT_EQ(err, -ERANGE, "bpf_map_update_batch -ERANGE")) + goto out; + + for (cpu =3D 0; cpu < nr_cpus; cpu++) { + memset(values, 0, value_sz_total); + + /* clear values across all CPUs */ + count =3D entries; + batch_opts.elem_flags =3D BPF_F_ALL_CPUS; + err =3D bpf_map_update_batch(map_fd, keys, values, &count, &batch_opts); + if (!ASSERT_OK(err, "bpf_map_update_batch all_cpus")) + goto out; + + /* update values on specified CPU */ + for (i =3D 0; i < entries; i++) + values[i] =3D value; + + count =3D entries; + batch_opts.elem_flags =3D (u64)cpu << 32 | BPF_F_CPU; + err =3D bpf_map_update_batch(map_fd, keys, values, &count, &batch_opts); + if (!ASSERT_OK(err, "bpf_map_update_batch specified cpu")) + goto out; + + /* lookup values on specified CPU */ + batch =3D 0; + count =3D entries; + memset(values, 0, entries * value_sz); + err =3D bpf_map_lookup_batch(map_fd, NULL, &batch, keys, values, &count,= &batch_opts); + if (!ASSERT_TRUE(!err || err =3D=3D -ENOENT, "bpf_map_lookup_batch speci= fied cpu")) + goto out; + + for (i =3D 0; i < entries; i++) + if (!ASSERT_EQ(values[i], value, + "bpf_map_lookup_batch value on specified cpu")) + goto out; + + /* lookup values from all CPUs */ + batch =3D 0; + count =3D entries; + batch_opts.elem_flags =3D 0; + memset(values_percpu, 0, roundup(value_sz, 8) * nr_cpus * entries); + err =3D bpf_map_lookup_batch(map_fd, NULL, &batch, keys, values_percpu, = &count, + &batch_opts); + if (!ASSERT_TRUE(!err || err =3D=3D -ENOENT, "bpf_map_lookup_batch all_c= pus")) + goto out; + + for (i =3D 0; i < entries; i++) { + values_row =3D (void *) values_percpu + + roundup(value_sz, 8) * i * nr_cpus; + for (j =3D 0; j < nr_cpus; j++) { + v =3D *(u32 *) (values_row + roundup(value_sz, 8) * j); + if (!ASSERT_EQ(v, j !=3D cpu ? 0 : value, + "bpf_map_lookup_batch value all_cpus")) + goto out; + } + } + } + +out: + free(values_percpu); + free(values); +} + + +static void test_percpu_map_cpu_flag(enum bpf_map_type map_type) +{ + struct percpu_alloc_array *skel; + size_t key_sz =3D sizeof(int); + int *keys, nr_cpus, i, err; + struct bpf_map *map; + u32 max_entries; + + nr_cpus =3D libbpf_num_possible_cpus(); + if (!ASSERT_GT(nr_cpus, 0, "libbpf_num_possible_cpus")) + return; + + max_entries =3D nr_cpus + 1; + keys =3D calloc(max_entries, key_sz); + if (!ASSERT_OK_PTR(keys, "calloc keys")) + return; + + for (i =3D 0; i < max_entries; i++) + keys[i] =3D i; + + skel =3D percpu_alloc_array__open(); + if (!ASSERT_OK_PTR(skel, "percpu_alloc_array__open")) { + free(keys); + return; + } + + map =3D skel->maps.percpu; + bpf_map__set_type(map, map_type); + bpf_map__set_max_entries(map, max_entries); + + err =3D percpu_alloc_array__load(skel); + if (!ASSERT_OK(err, "test_percpu_alloc__load")) + goto out; + + test_percpu_map_op_cpu_flag(map, keys, key_sz, max_entries - 1, nr_cpus, = true); +out: + percpu_alloc_array__destroy(skel); + free(keys); +} + +static void test_percpu_array_cpu_flag(void) +{ + test_percpu_map_cpu_flag(BPF_MAP_TYPE_PERCPU_ARRAY); +} + +static void test_percpu_hash_cpu_flag(void) +{ + test_percpu_map_cpu_flag(BPF_MAP_TYPE_PERCPU_HASH); +} + +static void test_lru_percpu_hash_cpu_flag(void) +{ + test_percpu_map_cpu_flag(BPF_MAP_TYPE_LRU_PERCPU_HASH); +} + +static void test_percpu_cgroup_storage_cpu_flag(void) +{ + struct percpu_alloc_array *skel =3D NULL; + struct bpf_cgroup_storage_key key; + int cgroup, prog_fd, nr_cpus, err; + struct bpf_map *map; + + nr_cpus =3D libbpf_num_possible_cpus(); + if (!ASSERT_GT(nr_cpus, 0, "libbpf_num_possible_cpus")) + return; + + err =3D setup_cgroup_environment(); + if (!ASSERT_OK(err, "setup_cgroup_environment")) + return; + + cgroup =3D create_and_get_cgroup("/cg_percpu"); + if (!ASSERT_GE(cgroup, 0, "create_and_get_cgroup")) { + cleanup_cgroup_environment(); + return; + } + + err =3D join_cgroup("/cg_percpu"); + if (!ASSERT_OK(err, "join_cgroup")) + goto out; + + skel =3D percpu_alloc_array__open_and_load(); + if (!ASSERT_OK_PTR(skel, "percpu_alloc_array__open_and_load")) + goto out; + + prog_fd =3D bpf_program__fd(skel->progs.cgroup_egress); + err =3D bpf_prog_attach(prog_fd, cgroup, BPF_CGROUP_INET_EGRESS, 0); + if (!ASSERT_OK(err, "bpf_prog_attach")) + goto out; + + map =3D skel->maps.percpu_cgroup_storage; + err =3D bpf_map_get_next_key(bpf_map__fd(map), NULL, &key); + if (!ASSERT_OK(err, "bpf_map_get_next_key")) + goto out; + + test_percpu_map_op_cpu_flag(map, &key, sizeof(key), 1, nr_cpus, false); +out: + bpf_prog_detach2(-1, cgroup, BPF_CGROUP_INET_EGRESS); + close(cgroup); + cleanup_cgroup_environment(); + percpu_alloc_array__destroy(skel); +} + +static void test_map_op_cpu_flag(enum bpf_map_type map_type) +{ + u32 max_entries =3D 1, count =3D max_entries; + u64 flags, batch =3D 0, val =3D 0; + int err, map_fd, key =3D 0; + LIBBPF_OPTS(bpf_map_batch_opts, batch_opts); + + map_fd =3D bpf_map_create(map_type, "test_cpu_flag", sizeof(int), sizeof(= u64), max_entries, + NULL); + if (!ASSERT_GE(map_fd, 0, "bpf_map_create")) + return; + + flags =3D BPF_F_ALL_CPUS; + err =3D bpf_map_update_elem(map_fd, &key, &val, flags); + ASSERT_ERR(err, "bpf_map_update_elem all_cpus"); + + batch_opts.elem_flags =3D BPF_F_ALL_CPUS; + err =3D bpf_map_update_batch(map_fd, &key, &val, &count, &batch_opts); + ASSERT_ERR(err, "bpf_map_update_batch all_cpus"); + + flags =3D BPF_F_CPU; + err =3D bpf_map_lookup_elem_flags(map_fd, &key, &val, flags); + ASSERT_ERR(err, "bpf_map_lookup_elem_flags cpu"); + + batch_opts.elem_flags =3D BPF_F_CPU; + err =3D bpf_map_lookup_batch(map_fd, NULL, &batch, &key, &val, &count, &b= atch_opts); + ASSERT_ERR(err, "bpf_map_lookup_batch cpu"); + + close(map_fd); +} + +static void test_array_cpu_flag(void) +{ + test_map_op_cpu_flag(BPF_MAP_TYPE_ARRAY); +} + +static void test_hash_cpu_flag(void) +{ + test_map_op_cpu_flag(BPF_MAP_TYPE_HASH); +} + void test_percpu_alloc(void) { if (test__start_subtest("array")) @@ -125,4 +441,16 @@ void test_percpu_alloc(void) test_cgrp_local_storage(); if (test__start_subtest("failure_tests")) test_failure(); + if (test__start_subtest("cpu_flag_percpu_array")) + test_percpu_array_cpu_flag(); + if (test__start_subtest("cpu_flag_percpu_hash")) + test_percpu_hash_cpu_flag(); + if (test__start_subtest("cpu_flag_lru_percpu_hash")) + test_lru_percpu_hash_cpu_flag(); + if (test__start_subtest("cpu_flag_percpu_cgroup_storage")) + test_percpu_cgroup_storage_cpu_flag(); + if (test__start_subtest("cpu_flag_array")) + test_array_cpu_flag(); + if (test__start_subtest("cpu_flag_hash")) + test_hash_cpu_flag(); } diff --git a/tools/testing/selftests/bpf/progs/percpu_alloc_array.c b/tools= /testing/selftests/bpf/progs/percpu_alloc_array.c index 37c2d2608ec0..ed6a2a93d5a5 100644 --- a/tools/testing/selftests/bpf/progs/percpu_alloc_array.c +++ b/tools/testing/selftests/bpf/progs/percpu_alloc_array.c @@ -187,4 +187,36 @@ int BPF_PROG(test_array_map_10) return 0; } =20 +struct { + __uint(type, BPF_MAP_TYPE_PERCPU_ARRAY); + __uint(max_entries, 2); + __type(key, int); + __type(value, u32); +} percpu SEC(".maps"); + +SEC("?fentry/bpf_fentry_test1") +int BPF_PROG(test_percpu_array, int x) +{ + u64 value =3D 0xDEADC0DE; + int key =3D 0; + + bpf_map_update_elem(&percpu, &key, &value, BPF_ANY); + return 0; +} + +struct { + __uint(type, BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE); + __type(key, struct bpf_cgroup_storage_key); + __type(value, u32); +} percpu_cgroup_storage SEC(".maps"); + +SEC("cgroup_skb/egress") +int cgroup_egress(struct __sk_buff *skb) +{ + u32 *val =3D bpf_get_local_storage(&percpu_cgroup_storage, 0); + + *val =3D 1; + return 1; +} + char _license[] SEC("license") =3D "GPL"; --=20 2.51.2