Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753058AbcJDEmB (ORCPT ); Tue, 4 Oct 2016 00:42:01 -0400 Received: from mail-pa0-f48.google.com ([209.85.220.48]:33744 "EHLO mail-pa0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752077AbcJDElg (ORCPT ); Tue, 4 Oct 2016 00:41:36 -0400 From: John Stultz To: lkml Cc: Colin Cross , Tejun Heo , Li Zefan , Jonathan Corbet , cgroups@vger.kernel.org, Android Kernel Team , Rom Lemarchand , Dmitry Shmidt , Todd Kjos , Christian Poetzsch , Amit Pundir , John Stultz Subject: [PATCH 1/2] cgroup: Add generic cgroup subsystem permission checks Date: Mon, 3 Oct 2016 21:41:29 -0700 Message-Id: <1475556090-6278-2-git-send-email-john.stultz@linaro.org> X-Mailer: git-send-email 1.9.1 In-Reply-To: <1475556090-6278-1-git-send-email-john.stultz@linaro.org> References: <1475556090-6278-1-git-send-email-john.stultz@linaro.org> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4587 Lines: 129 From: Colin Cross Rather than using explicit euid == 0 checks when trying to move tasks into a cgroup, move permission checks into each specific cgroup subsystem. If a subsystem does not specify a 'allow_attach' handler, then we fall back to doing the checks the old way. This patch adds a 'allow_attach' handler instead of reusing the 'can_attach' handler, since if the 'can_attach' handler was reused, a new cgroup that implements 'can_attach' but not the permission checks could end up with no permission checks at all. This also includes folded down fixes from: Christian Poetzsch Amit Pundir Cc: Tejun Heo Cc: Li Zefan Cc: Jonathan Corbet Cc: cgroups@vger.kernel.org Cc: Android Kernel Team Cc: Rom Lemarchand Cc: Colin Cross Cc: Dmitry Shmidt Cc: Todd Kjos Cc: Christian Poetzsch Cc: Amit Pundir Original-Author: San Mehat Signed-off-by: Colin Cross [jstultz: Rewording of commit message, folded down fixes] Signed-off-by: John Stultz --- Documentation/cgroup-v1/cgroups.txt | 9 +++++++++ include/linux/cgroup-defs.h | 1 + kernel/cgroup.c | 40 +++++++++++++++++++++++++++++++++++-- 3 files changed, 48 insertions(+), 2 deletions(-) diff --git a/Documentation/cgroup-v1/cgroups.txt b/Documentation/cgroup-v1/cgroups.txt index 308e5ff..295f026 100644 --- a/Documentation/cgroup-v1/cgroups.txt +++ b/Documentation/cgroup-v1/cgroups.txt @@ -578,6 +578,15 @@ is completely unused; @cgrp->parent is still valid. (Note - can also be called for a newly-created cgroup if an error occurs after this subsystem's create() method has been called for the new cgroup). +int allow_attach(struct cgroup *cgrp, struct cgroup_taskset *tset) +(cgroup_mutex held by caller) + +Called prior to moving a task into a cgroup; if the subsystem +returns an error, this will abort the attach operation. Used +to extend the permission checks - if all subsystems in a cgroup +return 0, the attach will be allowed to proceed, even if the +default permission check (root or same user) fails. + int can_attach(struct cgroup *cgrp, struct cgroup_taskset *tset) (cgroup_mutex held by caller) diff --git a/include/linux/cgroup-defs.h b/include/linux/cgroup-defs.h index 5b17de6..0f4548c 100644 --- a/include/linux/cgroup-defs.h +++ b/include/linux/cgroup-defs.h @@ -441,6 +441,7 @@ struct cgroup_subsys { void (*css_free)(struct cgroup_subsys_state *css); void (*css_reset)(struct cgroup_subsys_state *css); + int (*allow_attach)(struct cgroup_taskset *tset); int (*can_attach)(struct cgroup_taskset *tset); void (*cancel_attach)(struct cgroup_taskset *tset); void (*attach)(struct cgroup_taskset *tset); diff --git a/kernel/cgroup.c b/kernel/cgroup.c index d6b729b..e6afe2d 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -2833,6 +2833,25 @@ static int cgroup_attach_task(struct cgroup *dst_cgrp, return ret; } +static int cgroup_allow_attach(struct cgroup *cgrp, struct cgroup_taskset *tset) +{ + struct cgroup_subsys_state *css; + int i; + int ret; + + for_each_css(css, i, cgrp) { + if (css->ss->allow_attach) { + ret = css->ss->allow_attach(tset); + if (ret) + return ret; + } else { + return -EACCES; + } + } + + return 0; +} + static int cgroup_procs_write_permission(struct task_struct *task, struct cgroup *dst_cgrp, struct kernfs_open_file *of) @@ -2847,8 +2866,25 @@ static int cgroup_procs_write_permission(struct task_struct *task, */ if (!uid_eq(cred->euid, GLOBAL_ROOT_UID) && !uid_eq(cred->euid, tcred->uid) && - !uid_eq(cred->euid, tcred->suid)) - ret = -EACCES; + !uid_eq(cred->euid, tcred->suid)) { + /* + * if the default permission check fails, give each + * cgroup a chance to extend the permission check + */ + struct cgroup_taskset tset = { + .src_csets = LIST_HEAD_INIT(tset.src_csets), + .dst_csets = LIST_HEAD_INIT(tset.dst_csets), + .csets = &tset.src_csets, + }; + struct css_set *cset; + + cset = task_css_set(task); + list_add(&cset->mg_node, &tset.src_csets); + ret = cgroup_allow_attach(dst_cgrp, &tset); + list_del(&tset.src_csets); + if (ret) + ret = -EACCES; + } if (!ret && cgroup_on_dfl(dst_cgrp)) { struct super_block *sb = of->file->f_path.dentry->d_sb; -- 1.9.1