Received: from mail-qv1-f45.google.com (mail-qv1-f45.google.com [209.85.219.45]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 08C321925BC for ; Sun, 22 Mar 2026 00:45:05 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.219.45 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1774140307; cv=none; b=Ft44bOJn/h9y0tJu5M0KWm+95KJ/F7rRBrFaIz44Sh+LmDUFsgd4uY51uS1+1Aur/MWaKTifhzKbOIMvlz5XqBu1bzyOyaIrz4n8+3jW3MvAtyNEZg43TaQGSK9bFbv8WqseSiVlfz5EKou0+rZLRmyZERedCy+eu0cSid/iAio= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1774140307; c=relaxed/simple; bh=oW4mYUz+jXj7OkLf9uZLF7gFI7iw8sov9pc50u8Z+7U=; h=Mime-Version:Content-Type:Date:Message-Id:Cc:Subject:From:To: References:In-Reply-To; b=OdZcnmxPz+5hQfDVQ2PQ+hShPc5Z2uId1w+EwkXNdw7/zPkPeU6HggFm7x37KRVMMKJm/NjFFJJT3Q+x4YZlFjfVA8TWKALwG5LyYFAUZ0uVt4vqZ+dU78VnGo+LF++FQyZPwuQDoJYrWq7nvxS7t1Lb8R6AcDYBynj05MFofLo= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=etsalapatis.com; spf=pass smtp.mailfrom=etsalapatis.com; dkim=pass (2048-bit key) header.d=etsalapatis-com.20230601.gappssmtp.com header.i=@etsalapatis-com.20230601.gappssmtp.com header.b=ai+q8ie8; arc=none smtp.client-ip=209.85.219.45 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=etsalapatis.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=etsalapatis.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=etsalapatis-com.20230601.gappssmtp.com header.i=@etsalapatis-com.20230601.gappssmtp.com header.b="ai+q8ie8" Received: by mail-qv1-f45.google.com with SMTP id 6a1803df08f44-899ee87355dso36823636d6.1 for ; Sat, 21 Mar 2026 17:45:05 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=etsalapatis-com.20230601.gappssmtp.com; s=20230601; t=1774140305; x=1774745105; darn=vger.kernel.org; h=in-reply-to:references:to:from:subject:cc:message-id:date :content-transfer-encoding:mime-version:from:to:cc:subject:date :message-id:reply-to; bh=ub6j//PDVplmP1015QItMkzgMkrEL2KZLlzU1ilcsRs=; b=ai+q8ie8fw+NKa2WVFyeCsNsmZd67CvFc8zDhaTWwy/rdOaKL2sO5lCvWyKpCWq2BS wRN5nss0b/TNREe0Nr5zNW9nR3iM8o0k1z55z73siLQE6+n9OiNIZQ4Ejxs/d9Js8wEN W0KBcp6NpbPjxxN1ZKyqoZvJzbapG6CxnNg3h/EXpmNUS8U5zG/+5/nlp0lBQkZMWhSn jGruhNIOMexka53hhU96NoqByFbtP068B5fUFfmX1xFdRDXuvILySS/BygDzXbbSWkdT kjBerI2ra4Ny81kMaoFJ22aGPRu7c9uIYG2Ire4gRCbWxFrWzcSBGn5qp9fYbxWXOAE3 UXHQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1774140305; x=1774745105; h=in-reply-to:references:to:from:subject:cc:message-id:date :content-transfer-encoding:mime-version:x-gm-gg:x-gm-message-state :from:to:cc:subject:date:message-id:reply-to; bh=ub6j//PDVplmP1015QItMkzgMkrEL2KZLlzU1ilcsRs=; b=CAwbNL4XXLsjNju8XQOclRE5eTNG3ooqJVvGP6osJTLZa2aGzvVcWX+GkSQQuMcdLE SH1iYvJjLHEmPjiidQuTisAzk7L/zLy2cmD4n5BkR3+g7gW3SctSe+LIRDToVk8BVA3d P27xBjqeCZNSjO8A6J7Iz7nqrHsU4IbttEd/ZOgNlFv5P5DG5nhc3QANjjEQlg6At/45 CSsHF/OuRtIzJQbQE1Qdril+kgBEwxd6jv4KDsynIpYbEoAPZqdvXgnEL3uPGN8qtamf 9sm85mCx7e6qcxS42Kaf/InElYBwBTdfontgu7TfHIytCLziu2SZZapO/EHqp8e982yo 2KRA== X-Gm-Message-State: AOJu0YxEtpWEdHXxb/Jk0jnwqPBiHDSg00I3oW4K2+Y6fJl7LSlxft7m Sdbdz3Y5I1XSCb1h/oBWd8kBpxu1ubt2/q1rH2pTYBC7MBEXolJ/gxfcLJV62ysdM7XwJPUiH47 cmbkZZAk= X-Gm-Gg: ATEYQzwdNfVDy5baK1MUd1pf58bYXocLmfnRRJ/oVvLLTbPdeMEg3SbvDOGPkSSazeF Fiiho2VXh+197RxBc6vv6U9aZFMx830RgC3qTT/AYKLdS9FPTn/gYyCOQBg4Pm2pfTwqr8Frcl5 xwJpndyN94XQ/hq0hbLr++KU85qbKvaPr2oZBnbE3HGIzj2aSrn6e37k1g/YCyB3wendY1nXKaX QGH9lQuOpbG75EcjbJK0NLeE12n6E8Ea+Vh/Or0+4djxrJ4mGlSO7Wx4t0aY4silBVQtLkQWQ+D deYWLRkllr8yRWeuaVpwKzgXqqjzSUoTJUn8+8yfE0mm7M78ampISg42qtd2cG0DNvF/AKbkWR/ dcppo/XSBiaDfOY8fSiSITIewHLbSeS/t43nDaJwZAXPHXNlQ2WatLIuhbG+p0gf2OJnkdDV/aD F1S8OuOZdMqfpmJl2Sq39xn40= X-Received: by 2002:ad4:5968:0:b0:89a:110f:8946 with SMTP id 6a1803df08f44-89c859b5280mr136052986d6.13.1774140304947; Sat, 21 Mar 2026 17:45:04 -0700 (PDT) Received: from localhost ([140.174.219.137]) by smtp.gmail.com with ESMTPSA id 6a1803df08f44-89c85233ea6sm53211956d6.11.2026.03.21.17.45.04 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Sat, 21 Mar 2026 17:45:04 -0700 (PDT) Precedence: bulk X-Mailing-List: bpf@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=UTF-8 Date: Sat, 21 Mar 2026 20:45:02 -0400 Message-Id: Cc: , Subject: Re: [PATCH bpf-next v8 5/8] bpf: Add bpf_list_add_impl to insert node after a given list node From: "Emil Tsalapatis" To: "Chengkaitao" , , , , , , , , , , , , , , , X-Mailer: aerc 0.20.1 References: <20260316112843.78657-1-pilgrimtao@gmail.com> <20260316112843.78657-6-pilgrimtao@gmail.com> In-Reply-To: <20260316112843.78657-6-pilgrimtao@gmail.com> On Mon Mar 16, 2026 at 7:28 AM EDT, Chengkaitao wrote: > From: Kaitao Cheng > > Add a new kfunc bpf_list_add_impl(head, new, prev, meta, off) that > inserts 'new' after 'prev' in the BPF linked list. Both must be in > the same list; 'prev' must already be in the list. The new node must > be an owning reference (e.g. from bpf_obj_new); the kfunc consumes > that reference and the node becomes non-owning once inserted. > > We have added an additional parameter bpf_list_head *head to > bpf_list_add_impl, as the verifier requires the head parameter to > check whether the lock is being held. > > Returns 0 on success, -EINVAL if 'prev' is not in a list or 'new' > is already in a list (or duplicate insertion). On failure, the > kernel drops the passed-in node. > > Signed-off-by: Kaitao Cheng > --- > kernel/bpf/helpers.c | 14 ++++++++++++++ > kernel/bpf/verifier.c | 13 +++++++++++-- > 2 files changed, 25 insertions(+), 2 deletions(-) > > diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c > index a9665f97b3bc..dc4f8b4eec01 100644 > --- a/kernel/bpf/helpers.c > +++ b/kernel/bpf/helpers.c > @@ -2438,6 +2438,19 @@ __bpf_kfunc int bpf_list_push_back_impl(struct bpf= _list_head *head, > return __bpf_list_add(new, head, &h->prev, meta ? meta->record : NULL, = off); > } > =20 > +__bpf_kfunc int bpf_list_add_impl(struct bpf_list_head *head, > + struct bpf_list_node *new, > + struct bpf_list_node *prev, > + void *meta__ign, u64 off) > +{ > + struct bpf_list_node_kern *kn =3D (void *)new, *kp =3D (void *)prev; > + struct btf_struct_meta *meta =3D meta__ign; > + struct list_head *prev_ptr =3D &kp->list_head; > + You need to add a version of the check you added on patch 4 here to make sure the node owners are sane. The check was a WARN_ON because before the callers to __bpf_list_add didn't take a separate head and node argument from the program, so any inconsistency was a bug. With this helper this is no longer the case. Also check the comment in Sashiko. Can you either add a test that ensures the use-after-free is described is not the case, or otherwise change the error handling/verifier behavior to avoid it? I suspect it's an actual issue. In general your code is conflating two different issues in its error handling: 1) A node is being inserted into two different lists at the=20 same time, 2) the user has made a programming error and is supplying=20 an invalid head/prev combination. These are two very different issues, and 2) is only possible with this new kfunc. Can you add error handling for it directly in bpf_list_add_impl, and assume the head/prev combination is alsways valid in __bpf_list_add? > + return __bpf_list_add(kn, head, &prev_ptr, > + meta ? meta->record : NULL, off); > +} > + > static struct bpf_list_node *__bpf_list_del(struct bpf_list_head *head, > struct list_head *n) > { > @@ -4574,6 +4587,7 @@ BTF_ID_FLAGS(func, bpf_list_pop_back, KF_ACQUIRE | = KF_RET_NULL) > BTF_ID_FLAGS(func, bpf_list_del, KF_ACQUIRE | KF_RET_NULL) > BTF_ID_FLAGS(func, bpf_list_front, KF_RET_NULL) > BTF_ID_FLAGS(func, bpf_list_back, KF_RET_NULL) > +BTF_ID_FLAGS(func, bpf_list_add_impl) > BTF_ID_FLAGS(func, bpf_task_acquire, KF_ACQUIRE | KF_RCU | KF_RET_NULL) > BTF_ID_FLAGS(func, bpf_task_release, KF_RELEASE) > BTF_ID_FLAGS(func, bpf_rbtree_remove, KF_ACQUIRE | KF_RET_NULL) > diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c > index e928ad4290c7..98ddb370feb5 100644 > --- a/kernel/bpf/verifier.c > +++ b/kernel/bpf/verifier.c > @@ -12506,6 +12506,7 @@ enum special_kfunc_type { > KF_bpf_refcount_acquire_impl, > KF_bpf_list_push_front_impl, > KF_bpf_list_push_back_impl, > + KF_bpf_list_add_impl, > KF_bpf_list_pop_front, > KF_bpf_list_pop_back, > KF_bpf_list_del, > @@ -12567,6 +12568,7 @@ BTF_ID(func, bpf_obj_drop_impl) > BTF_ID(func, bpf_refcount_acquire_impl) > BTF_ID(func, bpf_list_push_front_impl) > BTF_ID(func, bpf_list_push_back_impl) > +BTF_ID(func, bpf_list_add_impl) > BTF_ID(func, bpf_list_pop_front) > BTF_ID(func, bpf_list_pop_back) > BTF_ID(func, bpf_list_del) > @@ -12644,6 +12646,7 @@ BTF_ID(func, bpf_stream_print_stack) > static const enum special_kfunc_type bpf_list_api_kfuncs[] =3D { > KF_bpf_list_push_front_impl, > KF_bpf_list_push_back_impl, > + KF_bpf_list_add_impl, > KF_bpf_list_pop_front, > KF_bpf_list_pop_back, > KF_bpf_list_del, > @@ -12655,6 +12658,7 @@ static const enum special_kfunc_type bpf_list_api= _kfuncs[] =3D { > static const enum special_kfunc_type bpf_list_node_api_kfuncs[] =3D { > KF_bpf_list_push_front_impl, > KF_bpf_list_push_back_impl, > + KF_bpf_list_add_impl, > KF_bpf_list_del, > }; > =20 > @@ -14345,6 +14349,7 @@ static int check_kfunc_call(struct bpf_verifier_e= nv *env, struct bpf_insn *insn, > =20 > if (meta.func_id =3D=3D special_kfunc_list[KF_bpf_list_push_front_impl]= || > meta.func_id =3D=3D special_kfunc_list[KF_bpf_list_push_back_impl] = || > + meta.func_id =3D=3D special_kfunc_list[KF_bpf_list_add_impl] || > meta.func_id =3D=3D special_kfunc_list[KF_bpf_rbtree_add_impl]) { Can you use the macros you defined on patch one to replace this list and the one below? > release_ref_obj_id =3D regs[BPF_REG_2].ref_obj_id; > insn_aux->insert_off =3D regs[BPF_REG_2].var_off.value; > @@ -23357,13 +23362,17 @@ static int fixup_kfunc_call(struct bpf_verifier= _env *env, struct bpf_insn *insn, > *cnt =3D 3; > } else if (desc->func_id =3D=3D special_kfunc_list[KF_bpf_list_push_bac= k_impl] || > desc->func_id =3D=3D special_kfunc_list[KF_bpf_list_push_front_impl= ] || > + desc->func_id =3D=3D special_kfunc_list[KF_bpf_list_add_impl] || > desc->func_id =3D=3D special_kfunc_list[KF_bpf_rbtree_add_impl]) { > struct btf_struct_meta *kptr_struct_meta =3D env->insn_aux_data[insn_i= dx].kptr_struct_meta; > int struct_meta_reg =3D BPF_REG_3; > int node_offset_reg =3D BPF_REG_4; > =20 > - /* rbtree_add has extra 'less' arg, so args-to-fixup are in diff regs = */ > - if (desc->func_id =3D=3D special_kfunc_list[KF_bpf_rbtree_add_impl]) { > + /* list/rbtree_add_impl have an extra arg (prev/less), > + * so args-to-fixup are in different regs. > + */ > + if (desc->func_id =3D=3D special_kfunc_list[KF_bpf_list_add_impl] || > + desc->func_id =3D=3D special_kfunc_list[KF_bpf_rbtree_add_impl]) { > struct_meta_reg =3D BPF_REG_4; > node_offset_reg =3D BPF_REG_5; > }