From patchwork Fri Oct 28 16:54:21 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Roberto Sassu X-Patchwork-Id: 13024080 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id D4001FA3742 for ; Fri, 28 Oct 2022 16:56:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230385AbiJ1Q4H (ORCPT ); Fri, 28 Oct 2022 12:56:07 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59316 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229631AbiJ1QzZ (ORCPT ); Fri, 28 Oct 2022 12:55:25 -0400 Received: from frasgout12.his.huawei.com (frasgout12.his.huawei.com [14.137.139.154]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 5584924940; Fri, 28 Oct 2022 09:55:16 -0700 (PDT) Received: from mail02.huawei.com (unknown [172.18.147.228]) by frasgout12.his.huawei.com (SkyGuard) with ESMTP id 4MzT6W0CC0z9xtVV; Sat, 29 Oct 2022 00:49:39 +0800 (CST) Received: from huaweicloud.com (unknown [10.204.63.22]) by APP1 (Coremail) with SMTP id LxC2BwBX9XFOCVxj63ccAA--.45991S2; Fri, 28 Oct 2022 17:54:51 +0100 (CET) From: Roberto Sassu To: ast@kernel.org, daniel@iogearbox.net, andrii@kernel.org, martin.lau@linux.dev, song@kernel.org, yhs@fb.com, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@google.com, haoluo@google.com, jolsa@kernel.org, mykolal@fb.com, revest@chromium.org, jackmanb@chromium.org, shuah@kernel.org, paul@paul-moore.com, casey@schaufler-ca.com, zohar@linux.ibm.com Cc: bpf@vger.kernel.org, linux-security-module@vger.kernel.org, linux-integrity@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org, nicolas.bouchinet@clip-os.org Subject: [RESEND][RFC][PATCH 1/3] lsm: Clarify documentation of vm_enough_memory hook Date: Fri, 28 Oct 2022 18:54:21 +0200 Message-Id: <20221028165423.386151-1-roberto.sassu@huaweicloud.com> X-Mailer: git-send-email 2.25.1 MIME-Version: 1.0 X-CM-TRANSID: LxC2BwBX9XFOCVxj63ccAA--.45991S2 X-Coremail-Antispam: 1UD129KBjvdXoWruFy8GF18ZrWxtw17KrWUCFg_yoWkXFb_u3 4fG348Xw4fXF4xKa1Ikr93tryrK3yfXr1qgF1Yq39IqFWDJas5Gw4xWFnxX3WDWwn293s5 uFyktrWxAwnIgjkaLaAFLSUrUUUUbb8apTn2vfkv8UJUUUU8Yxn0WfASr-VFAUDa7-sFnT 9fnUUIcSsGvfJTRUUUbs8YFVCjjxCrM7AC8VAFwI0_Xr0_Wr1l1xkIjI8I6I8E6xAIw20E Y4v20xvaj40_Wr0E3s1l1IIY67AEw4v_Jr0_Jr4l8cAvFVAK0II2c7xJM28CjxkF64kEwV A0rcxSw2x7M28EF7xvwVC0I7IYx2IY67AKxVWUJVWUCwA2z4x0Y4vE2Ix0cI8IcVCY1x02 67AKxVWxJVW8Jr1l84ACjcxK6I8E87Iv67AKxVW8JVWxJwA2z4x0Y4vEx4A2jsIEc7CjxV AFwI0_Gr1j6F4UJwAS0I0E0xvYzxvE52x082IY62kv0487Mc02F40EFcxC0VAKzVAqx4xG 6I80ewAv7VC0I7IYx2IY67AKxVWUGVWUXwAv7VC2z280aVAFwI0_Jr0_Gr1lOx8S6xCaFV Cjc4AY6r1j6r4UM4x0Y48IcxkI7VAKI48JM4IIrI8v6xkF7I0E8cxan2IY04v7MxkF7I0E n4kS14v26rWY6Fy7MxAIw28IcxkI7VAKI48JMxC20s026xCaFVCjc4AY6r1j6r4UMI8I3I 0E5I8CrVAFwI0_Jr0_Jr4lx2IqxVCjr7xvwVAFwI0_JrI_JrWlx4CE17CEb7AF67AKxVW8 ZVWrXwCIc40Y0x0EwIxGrwCI42IY6xIIjxv20xvE14v26r1j6r1xMIIF0xvE2Ix0cI8IcV CY1x0267AKxVWxJVW8Jr1lIxAIcVCF04k26cxKx2IYs7xG6Fyj6rWUJwCI42IY6I8E87Iv 67AKxVWUJVW8JwCI42IY6I8E87Iv6xkF7I0E14v26r4UJVWxJrUvcSsGvfC2KfnxnUUI43 ZEXa7IU0l1v3UUUUU== X-CM-SenderInfo: purev21wro2thvvxqx5xdzvxpfor3voofrz/1tbiAgATBF1jj4DNhQABst X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org From: Roberto Sassu include/linux/lsm_hooks.h reports the result of the LSM infrastructure to the callers, not what LSMs should return to the LSM infrastructure. Clarify that and add that returning 1 from the LSMs means calling __vm_enough_memory() with cap_sys_admin set, 0 without. Signed-off-by: Roberto Sassu Reviewed-by: KP Singh --- include/linux/lsm_hooks.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h index 4ec80b96c22e..f40b82ca91e7 100644 --- a/include/linux/lsm_hooks.h +++ b/include/linux/lsm_hooks.h @@ -1411,7 +1411,9 @@ * Check permissions for allocating a new virtual mapping. * @mm contains the mm struct it is being added to. * @pages contains the number of pages. - * Return 0 if permission is granted. + * Return 0 if permission is granted by LSMs to the caller. LSMs should + * return 1 if __vm_enough_memory() should be called with + * cap_sys_admin set, 0 if not. * * @ismaclabel: * Check if the extended attribute specified by @name From patchwork Fri Oct 28 16:54:22 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Roberto Sassu X-Patchwork-Id: 13024081 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B403DFA3746 for ; Fri, 28 Oct 2022 16:56:11 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230294AbiJ1Q4J (ORCPT ); Fri, 28 Oct 2022 12:56:09 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33248 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230297AbiJ1Qze (ORCPT ); Fri, 28 Oct 2022 12:55:34 -0400 Received: from frasgout12.his.huawei.com (frasgout12.his.huawei.com [14.137.139.154]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id A97931C4924; Fri, 28 Oct 2022 09:55:27 -0700 (PDT) Received: from mail02.huawei.com (unknown [172.18.147.227]) by frasgout12.his.huawei.com (SkyGuard) with ESMTP id 4MzT6k49qPz9xtVb; Sat, 29 Oct 2022 00:49:50 +0800 (CST) Received: from huaweicloud.com (unknown [10.204.63.22]) by APP1 (Coremail) with SMTP id LxC2BwBX9XFOCVxj63ccAA--.45991S3; Fri, 28 Oct 2022 17:55:03 +0100 (CET) From: Roberto Sassu To: ast@kernel.org, daniel@iogearbox.net, andrii@kernel.org, martin.lau@linux.dev, song@kernel.org, yhs@fb.com, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@google.com, haoluo@google.com, jolsa@kernel.org, mykolal@fb.com, revest@chromium.org, jackmanb@chromium.org, shuah@kernel.org, paul@paul-moore.com, casey@schaufler-ca.com, zohar@linux.ibm.com Cc: bpf@vger.kernel.org, linux-security-module@vger.kernel.org, linux-integrity@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org, nicolas.bouchinet@clip-os.org Subject: [RESEND][RFC][PATCH 2/3] bpf-lsm: Limit values that can be returned by security modules Date: Fri, 28 Oct 2022 18:54:22 +0200 Message-Id: <20221028165423.386151-2-roberto.sassu@huaweicloud.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20221028165423.386151-1-roberto.sassu@huaweicloud.com> References: <20221028165423.386151-1-roberto.sassu@huaweicloud.com> MIME-Version: 1.0 X-CM-TRANSID: LxC2BwBX9XFOCVxj63ccAA--.45991S3 X-Coremail-Antispam: 1UD129KBjvJXoW3Xw4rCr15ur1rGFWxAw4DXFb_yoW3Cw4Upr 4xJFyYkrsYvrZIqa4Iyan5Zws5AF1Fga1DKr1DGryIkrZ2vrykJw1UCryjqr9xWryUGrsa gr4qvF4Yg347ZaDanT9S1TB71UUUUUDqnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDU0xBIdaVrnRJUUUPqb4IE77IF4wAFF20E14v26rWj6s0DM7CY07I20VC2zVCF04k2 6cxKx2IYs7xG6rWj6s0DM7CIcVAFz4kK6r1j6r18M28IrcIa0xkI8VA2jI8067AKxVWUGw A2048vs2IY020Ec7CjxVAFwI0_Xr0E3s1l8cAvFVAK0II2c7xJM28CjxkF64kEwVA0rcxS w2x7M28EF7xvwVC0I7IYx2IY67AKxVWUJVWUCwA2z4x0Y4vE2Ix0cI8IcVCY1x0267AKxV W8Jr0_Cr1UM28EF7xvwVC2z280aVAFwI0_Gr0_Cr1l84ACjcxK6I8E87Iv6xkF7I0E14v2 6r4UJVWxJr1le2I262IYc4CY6c8Ij28IcVAaY2xG8wAqx4xG64xvF2IEw4CE5I8CrVC2j2 WlYx0E2Ix0cI8IcVAFwI0_JrI_JrylYx0Ex4A2jsIE14v26r1j6r4UMcvjeVCFs4IE7xkE bVWUJVW8JwACjcxG0xvY0x0EwIxGrwACI402YVCY1x02628vn2kIc2xKxwCY1x0262kKe7 AKxVWrXVW3AwCF04k20xvY0x0EwIxGrwCFx2IqxVCFs4IE7xkEbVWUJVW8JwC20s026c02 F40E14v26r1j6r18MI8I3I0E7480Y4vE14v26r106r1rMI8E67AF67kF1VAFwI0_GFv_Wr ylIxkGc2Ij64vIr41lIxAIcVC0I7IYx2IY67AKxVWUJVWUCwCI42IY6xIIjxv20xvEc7Cj xVAFwI0_Gr1j6F4UJwCI42IY6xAIw20EY4v20xvaj40_Jr0_JF4lIxAIcVC2z280aVAFwI 0_Jr0_Gr1lIxAIcVC2z280aVCY1x0267AKxVW8Jr0_Cr1UYxBIdaVFxhVjvjDU0xZFpf9x 07j-kuxUUUUU= X-CM-SenderInfo: purev21wro2thvvxqx5xdzvxpfor3voofrz/1tbiAQATBF1jj4TMWgAAs1 X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Roberto Sassu BPF LSM defines a bpf_lsm_*() function for each LSM hook, so that security modules can define their own implementation for the desired hooks. Unfortunately, BPF LSM does not restrict which values security modules can return (for non-void LSM hooks). Security modules might follow the conventions stated in include/linux/lsm_hooks.h, or put arbitrary values. This could cause big troubles, as the kernel is not ready to handle possibly malicious return values from LSMs. Until now, it was not the case, as each LSM is carefully reviewed and it won't be accepted if it does not meet the return value conventions. The biggest problem is when an LSM returns a positive value, instead of a negative value, as it could be converted to a pointer. Since such pointer escapes the IS_ERR() check, its use later in the code can cause unpredictable consequences (e.g. invalid memory access). Another problem is returning zero when an LSM is supposed to have done some operations. For example, the inode_init_security hook expects that their implementations return zero only if they set the name and value of the new xattr to be added to the new inode. Otherwise, other kernel subsystems might encounter unexpected conditions leading to a crash (e.g. evm_protected_xattr_common() getting NULL as argument). Finally, there are LSM hooks which are supposed to return just one as positive value, or non-negative values. Also in these cases, although it seems less critical, it is safer to return to callers of the LSM infrastructure more precisely what they expect. As eBPF allows code outside the kernel to run, it is its responsibility to ensure that only expected values are returned to LSM infrastructure callers. Create four new BTF ID sets, respectively for hooks that can return positive values, only one as positive value, that cannot return zero, and that cannot return negative values. Create also corresponding functions to check if the hook a security module is attached to belongs to one of the defined sets. Finally, check in the eBPF verifier the value returned by security modules for each attached LSM hook, and return -EINVAL (the security module cannot run) if the hook implementation does not satisfy the hook return value policy. Cc: stable@vger.kernel.org Fixes: 9d3fdea789c8 ("bpf: lsm: Provide attachment points for BPF LSM programs") Signed-off-by: Roberto Sassu --- include/linux/bpf_lsm.h | 24 ++++++++++++++++++ kernel/bpf/bpf_lsm.c | 56 +++++++++++++++++++++++++++++++++++++++++ kernel/bpf/verifier.c | 35 +++++++++++++++++++++++--- 3 files changed, 112 insertions(+), 3 deletions(-) diff --git a/include/linux/bpf_lsm.h b/include/linux/bpf_lsm.h index 4bcf76a9bb06..cd38aca4cfc0 100644 --- a/include/linux/bpf_lsm.h +++ b/include/linux/bpf_lsm.h @@ -28,6 +28,10 @@ int bpf_lsm_verify_prog(struct bpf_verifier_log *vlog, const struct bpf_prog *prog); bool bpf_lsm_is_sleepable_hook(u32 btf_id); +bool bpf_lsm_can_ret_pos_value(u32 btf_id); +bool bpf_lsm_can_ret_only_one_as_pos_value(u32 btf_id); +bool bpf_lsm_cannot_ret_zero(u32 btf_id); +bool bpf_lsm_cannot_ret_neg_value(u32 btf_id); static inline struct bpf_storage_blob *bpf_inode( const struct inode *inode) @@ -51,6 +55,26 @@ static inline bool bpf_lsm_is_sleepable_hook(u32 btf_id) return false; } +static inline bool bpf_lsm_can_ret_pos_value(u32 btf_id) +{ + return false; +} + +static inline bool bpf_lsm_can_ret_only_one_as_pos_value(u32 btf_id) +{ + return false; +} + +static inline bool bpf_lsm_cannot_ret_zero(u32 btf_id) +{ + return false; +} + +static inline bool bpf_lsm_cannot_ret_neg_value(u32 btf_id) +{ + return false; +} + static inline int bpf_lsm_verify_prog(struct bpf_verifier_log *vlog, const struct bpf_prog *prog) { diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c index d6c9b3705f24..3dcb70b2f978 100644 --- a/kernel/bpf/bpf_lsm.c +++ b/kernel/bpf/bpf_lsm.c @@ -348,6 +348,62 @@ bool bpf_lsm_is_sleepable_hook(u32 btf_id) return btf_id_set_contains(&sleepable_lsm_hooks, btf_id); } +/* The set of hooks which are allowed to return a positive value. */ +BTF_SET_START(pos_ret_value_lsm_hooks) +BTF_ID(func, bpf_lsm_vm_enough_memory) +BTF_ID(func, bpf_lsm_inode_getsecurity) +BTF_ID(func, bpf_lsm_inode_listsecurity) +BTF_ID(func, bpf_lsm_inode_need_killpriv) +BTF_ID(func, bpf_lsm_inode_copy_up_xattr) +BTF_ID(func, bpf_lsm_getprocattr) +BTF_ID(func, bpf_lsm_setprocattr) +BTF_ID(func, bpf_lsm_xfrm_state_pol_flow_match) +BTF_ID(func, bpf_lsm_key_getsecurity) +BTF_ID(func, bpf_lsm_ismaclabel) +BTF_ID(func, bpf_lsm_audit_rule_known) +BTF_ID(func, bpf_lsm_audit_rule_match) +BTF_SET_END(pos_ret_value_lsm_hooks) + +bool bpf_lsm_can_ret_pos_value(u32 btf_id) +{ + return btf_id_set_contains(&pos_ret_value_lsm_hooks, btf_id); +} + +BTF_SET_START(one_ret_value_lsm_hooks) +BTF_ID(func, bpf_lsm_vm_enough_memory) +BTF_ID(func, bpf_lsm_inode_copy_up_xattr) +BTF_ID(func, bpf_lsm_xfrm_state_pol_flow_match) +BTF_ID(func, bpf_lsm_ismaclabel) +BTF_ID(func, bpf_lsm_audit_rule_known) +BTF_ID(func, bpf_lsm_audit_rule_match) +BTF_SET_END(one_ret_value_lsm_hooks) + +bool bpf_lsm_can_ret_only_one_as_pos_value(u32 btf_id) +{ + return btf_id_set_contains(&one_ret_value_lsm_hooks, btf_id); +} + +/* The set of hooks which are not allowed to return zero. */ +BTF_SET_START(not_zero_ret_value_lsm_hooks) +BTF_ID(func, bpf_lsm_inode_init_security) +BTF_SET_END(not_zero_ret_value_lsm_hooks) + +bool bpf_lsm_cannot_ret_zero(u32 btf_id) +{ + return btf_id_set_contains(¬_zero_ret_value_lsm_hooks, btf_id); +} + +/* The set of hooks which are not allowed to return a negative value. */ +BTF_SET_START(not_neg_ret_value_lsm_hooks) +BTF_ID(func, bpf_lsm_vm_enough_memory) +BTF_ID(func, bpf_lsm_audit_rule_known) +BTF_SET_END(not_neg_ret_value_lsm_hooks) + +bool bpf_lsm_cannot_ret_neg_value(u32 btf_id) +{ + return btf_id_set_contains(¬_neg_ret_value_lsm_hooks, btf_id); +} + const struct bpf_prog_ops lsm_prog_ops = { }; diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 7f0a9f6cb889..099c1bf88fed 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -10623,9 +10623,38 @@ static int check_return_code(struct bpf_verifier_env *env) case BPF_PROG_TYPE_LSM: if (env->prog->expected_attach_type != BPF_LSM_CGROUP) { - /* Regular BPF_PROG_TYPE_LSM programs can return - * any value. - */ + /* < 0 */ + if (tnum_in(tnum_range((u64)(~0) << 31, (u64)(~0)), reg->var_off)) { + if (bpf_lsm_cannot_ret_neg_value(env->prog->aux->attach_btf_id)) { + verbose(env, "Invalid R0, cannot return negative value\n"); + return -EINVAL; + } + /* = 0 */ + } else if (tnum_equals_const(reg->var_off, 0)) { + if (bpf_lsm_cannot_ret_zero(env->prog->aux->attach_btf_id)) { + verbose(env, "Invalid R0, cannot return zero value\n"); + return -EINVAL; + } + /* = 1 */ + } else if (tnum_equals_const(reg->var_off, 1)) { + if (!bpf_lsm_can_ret_pos_value(env->prog->aux->attach_btf_id)) { + verbose(env, "Invalid R0, cannot return positive value\n"); + return -EINVAL; + } + /* > 1 */ + } else { + if (!bpf_lsm_can_ret_pos_value(env->prog->aux->attach_btf_id)) { + verbose(env, "Invalid R0, cannot return positive value\n"); + return -EINVAL; + } + + if (bpf_lsm_can_ret_only_one_as_pos_value(env->prog->aux->attach_btf_id)) { + verbose(env, + "Invalid R0, can return only one as positive value\n"); + return -EINVAL; + } + } + return 0; } if (!env->prog->aux->attach_func_proto->type) { From patchwork Fri Oct 28 16:54:23 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Roberto Sassu X-Patchwork-Id: 13024082 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id A3977FA3743 for ; Fri, 28 Oct 2022 16:56:38 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230354AbiJ1Q4g (ORCPT ); Fri, 28 Oct 2022 12:56:36 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:34370 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230366AbiJ1Qzq (ORCPT ); Fri, 28 Oct 2022 12:55:46 -0400 Received: from frasgout11.his.huawei.com (frasgout11.his.huawei.com [14.137.139.23]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4EEB01CCCD1; Fri, 28 Oct 2022 09:55:37 -0700 (PDT) Received: from mail02.huawei.com (unknown [172.18.147.228]) by frasgout11.his.huawei.com (SkyGuard) with ESMTP id 4MzT5x3dj4z9xFQb; Sat, 29 Oct 2022 00:49:09 +0800 (CST) Received: from huaweicloud.com (unknown [10.204.63.22]) by APP1 (Coremail) with SMTP id LxC2BwBX9XFOCVxj63ccAA--.45991S4; Fri, 28 Oct 2022 17:55:14 +0100 (CET) From: Roberto Sassu To: ast@kernel.org, daniel@iogearbox.net, andrii@kernel.org, martin.lau@linux.dev, song@kernel.org, yhs@fb.com, john.fastabend@gmail.com, kpsingh@kernel.org, sdf@google.com, haoluo@google.com, jolsa@kernel.org, mykolal@fb.com, revest@chromium.org, jackmanb@chromium.org, shuah@kernel.org, paul@paul-moore.com, casey@schaufler-ca.com, zohar@linux.ibm.com Cc: bpf@vger.kernel.org, linux-security-module@vger.kernel.org, linux-integrity@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org, nicolas.bouchinet@clip-os.org Subject: [RESEND][RFC][PATCH 3/3] selftests/bpf: Check if return values of LSM programs are allowed Date: Fri, 28 Oct 2022 18:54:23 +0200 Message-Id: <20221028165423.386151-3-roberto.sassu@huaweicloud.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20221028165423.386151-1-roberto.sassu@huaweicloud.com> References: <20221028165423.386151-1-roberto.sassu@huaweicloud.com> MIME-Version: 1.0 X-CM-TRANSID: LxC2BwBX9XFOCVxj63ccAA--.45991S4 X-Coremail-Antispam: 1UD129KBjvJXoWxJF1kAr4UAFyxAr48Cr17ZFb_yoWrZw1Dp3 WrZw1jkF40vF4avFWrK397uayS9FW7CrW5KwnxZwnrZa97JF4xW3W5tFy5Zr13Gr15Gr9Y qr17Can5u3WUZa7anT9S1TB71UUUUUDqnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDU0xBIdaVrnRJUUUPqb4IE77IF4wAFF20E14v26rWj6s0DM7CY07I20VC2zVCF04k2 6cxKx2IYs7xG6rWj6s0DM7CIcVAFz4kK6r1j6r18M28IrcIa0xkI8VA2jI8067AKxVWUXw A2048vs2IY020Ec7CjxVAFwI0_Xr0E3s1l8cAvFVAK0II2c7xJM28CjxkF64kEwVA0rcxS w2x7M28EF7xvwVC0I7IYx2IY67AKxVWUCVW8JwA2z4x0Y4vE2Ix0cI8IcVCY1x0267AKxV W8Jr0_Cr1UM28EF7xvwVC2z280aVAFwI0_Gr0_Cr1l84ACjcxK6I8E87Iv6xkF7I0E14v2 6r4UJVWxJr1le2I262IYc4CY6c8Ij28IcVAaY2xG8wAqx4xG64xvF2IEw4CE5I8CrVC2j2 WlYx0E2Ix0cI8IcVAFwI0_JrI_JrylYx0Ex4A2jsIE14v26r1j6r4UMcvjeVCFs4IE7xkE bVWUJVW8JwACjcxG0xvY0x0EwIxGrwACI402YVCY1x02628vn2kIc2xKxwCY1x0262kKe7 AKxVWrXVW3AwCF04k20xvY0x0EwIxGrwCFx2IqxVCFs4IE7xkEbVWUJVW8JwC20s026c02 F40E14v26r1j6r18MI8I3I0E7480Y4vE14v26r106r1rMI8E67AF67kF1VAFwI0_GFv_Wr ylIxkGc2Ij64vIr41lIxAIcVC0I7IYx2IY67AKxVWUCVW8JwCI42IY6xIIjxv20xvEc7Cj xVAFwI0_Gr1j6F4UJwCI42IY6xAIw20EY4v20xvaj40_Jr0_JF4lIxAIcVC2z280aVAFwI 0_Jr0_Gr1lIxAIcVC2z280aVCY1x0267AKxVW8Jr0_Cr1UYxBIdaVFxhVjvjDU0xZFpf9x 07jTbyZUUUUU= X-CM-SenderInfo: purev21wro2thvvxqx5xdzvxpfor3voofrz/1tbiAgATBF1jj4DNkwAAs6 X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net From: Roberto Sassu Ensure that the eBPF verifier allows to load only LSM programs that return an allowed value depending on the LSM hook they attach to. Signed-off-by: Roberto Sassu --- .../testing/selftests/bpf/verifier/lsm_ret.c | 148 ++++++++++++++++++ 1 file changed, 148 insertions(+) create mode 100644 tools/testing/selftests/bpf/verifier/lsm_ret.c diff --git a/tools/testing/selftests/bpf/verifier/lsm_ret.c b/tools/testing/selftests/bpf/verifier/lsm_ret.c new file mode 100644 index 000000000000..1a11f47fb24a --- /dev/null +++ b/tools/testing/selftests/bpf/verifier/lsm_ret.c @@ -0,0 +1,148 @@ +{ + "lsm return value: positive not allowed, return -EPERM", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, -EPERM), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "inode_permission", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: positive not allowed, return zero", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "inode_permission", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: positive not allowed, return one", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "inode_permission", + .expected_attach_type = BPF_LSM_MAC, + .errstr = "Invalid R0, cannot return positive value", + .result = REJECT, +}, +{ + "lsm return value: zero/positive not allowed, return -EPERM", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, -EPERM), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "inode_init_security", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: zero/positive not allowed, return zero", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "inode_init_security", + .expected_attach_type = BPF_LSM_MAC, + .errstr = "Invalid R0, cannot return zero value", + .result = REJECT, +}, +{ + "lsm return value: zero/positive not allowed, return one", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "inode_init_security", + .expected_attach_type = BPF_LSM_MAC, + .errstr = "Invalid R0, cannot return positive value", + .result = REJECT, +}, +{ + "lsm return value: positive allowed, return one", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "getprocattr", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: positive allowed, return two", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 2), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "getprocattr", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: only one allowed, return one", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "audit_rule_match", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: only one allowed, return two", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 2), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "audit_rule_match", + .expected_attach_type = BPF_LSM_MAC, + .errstr = "Invalid R0, can return only one as positive value", + .result = REJECT, +}, +{ + "lsm return value: negative not allowed, return -EPERM", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, -EPERM), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "vm_enough_memory", + .expected_attach_type = BPF_LSM_MAC, + .errstr = "Invalid R0, cannot return negative value", + .result = REJECT, +}, +{ + "lsm return value: negative not allowed, return zero", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 0), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "vm_enough_memory", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +}, +{ + "lsm return value: negative not allowed, return one", + .insns = { + BPF_MOV64_IMM(BPF_REG_0, 1), + BPF_EXIT_INSN(), + }, + .prog_type = BPF_PROG_TYPE_LSM, + .kfunc = "vm_enough_memory", + .expected_attach_type = BPF_LSM_MAC, + .result = ACCEPT, +},