From patchwork Mon Feb 13 15:32:25 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andreas Gruenbacher X-Patchwork-Id: 9570115 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 4FA3160572 for ; Mon, 13 Feb 2017 15:41:43 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 4546528173 for ; Mon, 13 Feb 2017 15:41:43 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 3A0A128355; Mon, 13 Feb 2017 15:41:43 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=unavailable version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 50A1628173 for ; Mon, 13 Feb 2017 15:41:42 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751456AbdBMPd0 (ORCPT ); Mon, 13 Feb 2017 10:33:26 -0500 Received: from mx1.redhat.com ([209.132.183.28]:45221 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753426AbdBMPdY (ORCPT ); Mon, 13 Feb 2017 10:33:24 -0500 Received: from int-mx11.intmail.prod.int.phx2.redhat.com (int-mx11.intmail.prod.int.phx2.redhat.com [10.5.11.24]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id AAD698124B; Mon, 13 Feb 2017 15:33:19 +0000 (UTC) Received: from nux.redhat.com (ovpn-117-141.ams2.redhat.com [10.36.117.141]) by int-mx11.intmail.prod.int.phx2.redhat.com (8.14.4/8.14.4) with ESMTP id v1DFWd4T006129; Mon, 13 Feb 2017 10:33:15 -0500 From: Andreas Gruenbacher To: Alexander Viro Cc: Andreas Gruenbacher , Christoph Hellwig , "Theodore Ts'o" , Andreas Dilger , "J. Bruce Fields" , Jeff Layton , Trond Myklebust , Anna Schumaker , Dave Chinner , linux-ext4@vger.kernel.org, linux-xfs@vger.kernel.org, linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-nfs@vger.kernel.org, linux-cifs@vger.kernel.org, linux-api@vger.kernel.org Subject: [PATCH v28 09/21] vfs: Cache base_acl objects in inodes Date: Mon, 13 Feb 2017 16:32:25 +0100 Message-Id: <1486999957-2381-10-git-send-email-agruenba@redhat.com> In-Reply-To: <1486999957-2381-1-git-send-email-agruenba@redhat.com> References: <1486999957-2381-1-git-send-email-agruenba@redhat.com> X-Scanned-By: MIMEDefang 2.68 on 10.5.11.24 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.25]); Mon, 13 Feb 2017 15:33:19 +0000 (UTC) Sender: linux-cifs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-cifs@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP POSIX ACLs and richacls are both objects allocated by kmalloc() with a reference count which are freed by kfree_rcu(). An inode can either cache an access and a default POSIX ACL, or a richacl (richacls don't have default acls). To allow an inode to cache either of the two kinds of acls, introduce a new base_acl type and convert i_acl and i_default_acl to that type. In most cases, the vfs doesn't care which kind of acl an inode caches (if any). Signed-off-by: Andreas Gruenbacher --- fs/9p/acl.c | 8 +++--- fs/f2fs/acl.c | 4 +-- fs/inode.c | 32 ++++++++++++++++++++-- fs/jffs2/acl.c | 6 +++-- fs/namei.c | 33 ++++++++++++----------- fs/nfs/nfs3acl.c | 14 +++++----- fs/posix_acl.c | 69 +++++++++++++---------------------------------- fs/richacl.c | 4 +-- include/linux/fs.h | 34 +++++++++++++++++++---- include/linux/posix_acl.h | 19 +++++++------ include/linux/richacl.h | 9 +++---- 11 files changed, 128 insertions(+), 104 deletions(-) diff --git a/fs/9p/acl.c b/fs/9p/acl.c index b3c2cc7..0e2c114 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -87,14 +87,14 @@ int v9fs_get_acl(struct inode *inode, struct p9_fid *fid) static struct posix_acl *v9fs_get_cached_acl(struct inode *inode, int type) { - struct posix_acl *acl; + struct base_acl *base_acl; /* * 9p Always cache the acl value when * instantiating the inode (v9fs_inode_from_fid) */ - acl = get_cached_acl(inode, type); - BUG_ON(is_uncached_acl(acl)); - return acl; + base_acl = get_cached_acl(inode, type); + BUG_ON(is_uncached_acl(base_acl)); + return posix_acl(base_acl); } struct posix_acl *v9fs_iop_get_acl(struct inode *inode, int type) diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c index 8f48769..8c8bd4e 100644 --- a/fs/f2fs/acl.c +++ b/fs/f2fs/acl.c @@ -266,7 +266,7 @@ static struct posix_acl *f2fs_acl_clone(const struct posix_acl *acl, sizeof(struct posix_acl_entry); clone = kmemdup(acl, size, flags); if (clone) - atomic_set(&clone->a_refcount, 1); + base_acl_init(&clone->a_base); } return clone; } @@ -278,8 +278,6 @@ static int f2fs_acl_create_masq(struct posix_acl *acl, umode_t *mode_p) umode_t mode = *mode_p; int not_equiv = 0; - /* assert(atomic_read(acl->a_refcount) == 1); */ - FOREACH_ACL_ENTRY(pa, acl, pe) { switch(pa->e_tag) { case ACL_USER_OBJ: diff --git a/fs/inode.c b/fs/inode.c index 88110fd..cd77ea1 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -242,14 +242,42 @@ void __destroy_inode(struct inode *inode) #ifdef CONFIG_FS_POSIX_ACL if (inode->i_acl && !is_uncached_acl(inode->i_acl)) - posix_acl_release(inode->i_acl); + base_acl_put(inode->i_acl); if (inode->i_default_acl && !is_uncached_acl(inode->i_default_acl)) - posix_acl_release(inode->i_default_acl); + base_acl_put(inode->i_default_acl); #endif this_cpu_dec(nr_inodes); } EXPORT_SYMBOL(__destroy_inode); +#ifdef CONFIG_FS_POSIX_ACL +struct base_acl *__get_cached_acl(struct base_acl **p) +{ + struct base_acl *base_acl; + + for (;;) { + rcu_read_lock(); + base_acl = rcu_dereference(*p); + if (!base_acl || is_uncached_acl(base_acl) || + atomic_inc_not_zero(&base_acl->ba_refcount)) + break; + rcu_read_unlock(); + cpu_relax(); + } + rcu_read_unlock(); + return base_acl; +} + +void __forget_cached_acl(struct base_acl **p) +{ + struct base_acl *old; + + old = xchg(p, ACL_NOT_CACHED); + if (!is_uncached_acl(old)) + base_acl_put(old); +} +#endif + static void i_callback(struct rcu_head *head) { struct inode *inode = container_of(head, struct inode, i_rcu); diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c index 7ebacf1..89940c4 100644 --- a/fs/jffs2/acl.c +++ b/fs/jffs2/acl.c @@ -291,13 +291,15 @@ int jffs2_init_acl_post(struct inode *inode) int rc; if (inode->i_default_acl) { - rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_DEFAULT, inode->i_default_acl); + rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_DEFAULT, + posix_acl(inode->i_default_acl)); if (rc) return rc; } if (inode->i_acl) { - rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_ACCESS, inode->i_acl); + rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_ACCESS, + posix_acl(inode->i_acl)); if (rc) return rc; } diff --git a/fs/namei.c b/fs/namei.c index ba91d8f..4b53754 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -260,25 +260,28 @@ void putname(struct filename *name) static int check_acl(struct inode *inode, int mask) { #ifdef CONFIG_FS_POSIX_ACL - struct posix_acl *acl; - if (mask & MAY_NOT_BLOCK) { - acl = get_cached_acl_rcu(inode, ACL_TYPE_ACCESS); - if (!acl) + struct base_acl *base_acl; + + base_acl = rcu_dereference(inode->i_acl); + if (!base_acl) return -EAGAIN; /* no ->get_acl() calls in RCU mode... */ - if (is_uncached_acl(acl)) + if (is_uncached_acl(base_acl)) return -ECHILD; - return posix_acl_permission(inode, acl, mask & ~MAY_NOT_BLOCK); - } - - acl = get_acl(inode, ACL_TYPE_ACCESS); - if (IS_ERR(acl)) - return PTR_ERR(acl); - if (acl) { - int error = posix_acl_permission(inode, acl, mask); - posix_acl_release(acl); - return error; + return posix_acl_permission(inode, posix_acl(base_acl), + mask & ~MAY_NOT_BLOCK); + } else { + struct posix_acl *acl; + + acl = get_acl(inode, ACL_TYPE_ACCESS); + if (IS_ERR(acl)) + return PTR_ERR(acl); + if (acl) { + int error = posix_acl_permission(inode, acl, mask); + posix_acl_release(acl); + return error; + } } #endif diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 720d92f5..2b70944 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -16,28 +16,28 @@ * caching get_acl results in a race-free way. See fs/posix_acl.c:get_acl() * for explanations. */ -static void nfs3_prepare_get_acl(struct posix_acl **p) +static void nfs3_prepare_get_acl(struct base_acl **p) { - struct posix_acl *sentinel = uncached_acl_sentinel(current); + struct base_acl *sentinel = uncached_acl_sentinel(current); if (cmpxchg(p, ACL_NOT_CACHED, sentinel) != ACL_NOT_CACHED) { /* Not the first reader or sentinel already in place. */ } } -static void nfs3_complete_get_acl(struct posix_acl **p, struct posix_acl *acl) +static void nfs3_complete_get_acl(struct base_acl **p, struct posix_acl *acl) { - struct posix_acl *sentinel = uncached_acl_sentinel(current); + struct base_acl *sentinel = uncached_acl_sentinel(current); /* Only cache the ACL if our sentinel is still in place. */ posix_acl_dup(acl); - if (cmpxchg(p, sentinel, acl) != sentinel) + if (cmpxchg(p, sentinel, &acl->a_base) != sentinel) posix_acl_release(acl); } -static void nfs3_abort_get_acl(struct posix_acl **p) +static void nfs3_abort_get_acl(struct base_acl **p) { - struct posix_acl *sentinel = uncached_acl_sentinel(current); + struct base_acl *sentinel = uncached_acl_sentinel(current); /* Remove our sentinel upon failure. */ cmpxchg(p, sentinel, ACL_NOT_CACHED); diff --git a/fs/posix_acl.c b/fs/posix_acl.c index c9d48dc..5d6c5fd 100644 --- a/fs/posix_acl.c +++ b/fs/posix_acl.c @@ -21,7 +21,7 @@ #include #include -static struct posix_acl **acl_by_type(struct inode *inode, int type) +static inline struct base_acl **acl_by_type(struct inode *inode, int type) { switch (type) { case ACL_TYPE_ACCESS: @@ -33,51 +33,23 @@ static struct posix_acl **acl_by_type(struct inode *inode, int type) } } -struct posix_acl *get_cached_acl(struct inode *inode, int type) +struct base_acl *get_cached_acl(struct inode *inode, int type) { - struct posix_acl **p = acl_by_type(inode, type); - struct posix_acl *acl; - - for (;;) { - rcu_read_lock(); - acl = rcu_dereference(*p); - if (!acl || is_uncached_acl(acl) || - atomic_inc_not_zero(&acl->a_refcount)) - break; - rcu_read_unlock(); - cpu_relax(); - } - rcu_read_unlock(); - return acl; + return __get_cached_acl(acl_by_type(inode, type)); } EXPORT_SYMBOL(get_cached_acl); -struct posix_acl *get_cached_acl_rcu(struct inode *inode, int type) -{ - return rcu_dereference(*acl_by_type(inode, type)); -} -EXPORT_SYMBOL(get_cached_acl_rcu); - void set_cached_acl(struct inode *inode, int type, struct posix_acl *acl) { - struct posix_acl **p = acl_by_type(inode, type); - struct posix_acl *old; + struct base_acl **p = acl_by_type(inode, type); + struct base_acl *old; - old = xchg(p, posix_acl_dup(acl)); + old = xchg(p, &posix_acl_dup(acl)->a_base); if (!is_uncached_acl(old)) - posix_acl_release(old); + base_acl_put(old); } EXPORT_SYMBOL(set_cached_acl); -static void __forget_cached_acl(struct posix_acl **p) -{ - struct posix_acl *old; - - old = xchg(p, ACL_NOT_CACHED); - if (!is_uncached_acl(old)) - posix_acl_release(old); -} - void forget_cached_acl(struct inode *inode, int type) { __forget_cached_acl(acl_by_type(inode, type)); @@ -93,25 +65,24 @@ EXPORT_SYMBOL(forget_all_cached_acls); struct posix_acl *get_acl(struct inode *inode, int type) { - void *sentinel; - struct posix_acl **p; + struct base_acl **p = acl_by_type(inode, type); + struct base_acl *sentinel, *base_acl; struct posix_acl *acl; + if (!IS_POSIXACL(inode)) + return NULL; + /* * The sentinel is used to detect when another operation like * set_cached_acl() or forget_cached_acl() races with get_acl(). * It is guaranteed that is_uncached_acl(sentinel) is true. */ - acl = get_cached_acl(inode, type); - if (!is_uncached_acl(acl)) - return acl; - - if (!IS_POSIXACL(inode)) - return NULL; + base_acl = __get_cached_acl(p); + if (!is_uncached_acl(base_acl)) + return posix_acl(base_acl); sentinel = uncached_acl_sentinel(current); - p = acl_by_type(inode, type); /* * If the ACL isn't being read yet, set our sentinel. Otherwise, the @@ -151,7 +122,7 @@ struct posix_acl *get_acl(struct inode *inode, int type) * Cache the result, but only if our sentinel is still in place. */ posix_acl_dup(acl); - if (unlikely(cmpxchg(p, sentinel, acl) != sentinel)) + if (unlikely(cmpxchg(p, sentinel, &acl->a_base) != sentinel)) posix_acl_release(acl); return acl; } @@ -163,7 +134,7 @@ EXPORT_SYMBOL(get_acl); void posix_acl_init(struct posix_acl *acl, int count) { - atomic_set(&acl->a_refcount, 1); + base_acl_init(&acl->a_base); acl->a_count = count; } EXPORT_SYMBOL(posix_acl_init); @@ -196,7 +167,7 @@ posix_acl_clone(const struct posix_acl *acl, gfp_t flags) sizeof(struct posix_acl_entry); clone = kmemdup(acl, size, flags); if (clone) - atomic_set(&clone->a_refcount, 1); + base_acl_init(&clone->a_base); } return clone; } @@ -418,8 +389,6 @@ static int posix_acl_create_masq(struct posix_acl *acl, umode_t *mode_p) umode_t mode = *mode_p; int not_equiv = 0; - /* assert(atomic_read(acl->a_refcount) == 1); */ - FOREACH_ACL_ENTRY(pa, acl, pe) { switch(pa->e_tag) { case ACL_USER_OBJ: @@ -473,8 +442,6 @@ static int __posix_acl_chmod_masq(struct posix_acl *acl, umode_t mode) struct posix_acl_entry *group_obj = NULL, *mask_obj = NULL; struct posix_acl_entry *pa, *pe; - /* assert(atomic_read(acl->a_refcount) == 1); */ - FOREACH_ACL_ENTRY(pa, acl, pe) { switch(pa->e_tag) { case ACL_USER_OBJ: diff --git a/fs/richacl.c b/fs/richacl.c index 0443090..57a4048 100644 --- a/fs/richacl.c +++ b/fs/richacl.c @@ -31,7 +31,7 @@ richacl_alloc(int count, gfp_t gfp) struct richacl *acl = kzalloc(size, gfp); if (acl) { - atomic_set(&acl->a_refcount, 1); + base_acl_init(&acl->a_base); acl->a_count = count; } return acl; @@ -50,7 +50,7 @@ richacl_clone(const struct richacl *acl, gfp_t gfp) if (dup) { memcpy(dup, acl, size); - atomic_set(&dup->a_refcount, 1); + base_acl_init(&dup->a_base); } return dup; } diff --git a/include/linux/fs.h b/include/linux/fs.h index 3de40d3..b33823c 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -530,18 +530,22 @@ static inline void mapping_allow_writable(struct address_space *mapping) #define i_size_ordered_init(inode) do { } while (0) #endif +struct base_acl { + atomic_t ba_refcount; + struct rcu_head ba_rcu; +}; struct posix_acl; #define ACL_NOT_CACHED ((void *)(-1)) #define ACL_DONT_CACHE ((void *)(-3)) -static inline struct posix_acl * +static inline struct base_acl * uncached_acl_sentinel(struct task_struct *task) { return (void *)task + 1; } static inline bool -is_uncached_acl(struct posix_acl *acl) +is_uncached_acl(struct base_acl *acl) { return (long)acl & 1; } @@ -564,9 +568,9 @@ struct inode { kgid_t i_gid; unsigned int i_flags; -#ifdef CONFIG_FS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; +#if defined(CONFIG_FS_POSIX_ACL) + struct base_acl *i_acl; + struct base_acl *i_default_acl; #endif const struct inode_operations *i_op; @@ -3184,4 +3188,24 @@ static inline bool dir_relax_shared(struct inode *inode) extern bool path_noexec(const struct path *path); extern void inode_nohighmem(struct inode *inode); +static inline void base_acl_get(struct base_acl *acl) +{ + if (acl) + atomic_inc(&acl->ba_refcount); +} + +static inline void base_acl_put(struct base_acl *acl) +{ + if (acl && atomic_dec_and_test(&acl->ba_refcount)) + kfree_rcu(acl, ba_rcu); +} + +static inline void base_acl_init(struct base_acl *acl) +{ + atomic_set(&acl->ba_refcount, 1); +} + +extern struct base_acl *__get_cached_acl(struct base_acl **); +extern void __forget_cached_acl(struct base_acl **); + #endif /* _LINUX_FS_H */ diff --git a/include/linux/posix_acl.h b/include/linux/posix_acl.h index 5a9a739..abfb786 100644 --- a/include/linux/posix_acl.h +++ b/include/linux/posix_acl.h @@ -23,8 +23,7 @@ struct posix_acl_entry { }; struct posix_acl { - atomic_t a_refcount; - struct rcu_head a_rcu; + struct base_acl a_base; /* must be first, see posix_acl_release() */ unsigned int a_count; struct posix_acl_entry a_entries[0]; }; @@ -39,8 +38,7 @@ struct posix_acl { static inline struct posix_acl * posix_acl_dup(struct posix_acl *acl) { - if (acl) - atomic_inc(&acl->a_refcount); + base_acl_get(&acl->a_base); return acl; } @@ -50,10 +48,16 @@ posix_acl_dup(struct posix_acl *acl) static inline void posix_acl_release(struct posix_acl *acl) { - if (acl && atomic_dec_and_test(&acl->a_refcount)) - kfree_rcu(acl, a_rcu); + BUILD_BUG_ON(offsetof(struct posix_acl, a_base) != 0); + base_acl_put(&acl->a_base); } +static inline struct posix_acl * +posix_acl(struct base_acl *base_acl) +{ + BUILD_BUG_ON(offsetof(struct posix_acl, a_base) != 0); + return container_of(base_acl, struct posix_acl, a_base); +} /* posix_acl.c */ @@ -78,8 +82,7 @@ extern int posix_acl_update_mode(struct inode *, umode_t *, struct posix_acl **) extern int simple_set_acl(struct inode *, struct posix_acl *, int); extern int simple_acl_create(struct inode *, struct inode *); -struct posix_acl *get_cached_acl(struct inode *inode, int type); -struct posix_acl *get_cached_acl_rcu(struct inode *inode, int type); +struct base_acl *get_cached_acl(struct inode *inode, int type); void set_cached_acl(struct inode *inode, int type, struct posix_acl *acl); void forget_cached_acl(struct inode *inode, int type); void forget_all_cached_acls(struct inode *inode); diff --git a/include/linux/richacl.h b/include/linux/richacl.h index 700888b..d535206 100644 --- a/include/linux/richacl.h +++ b/include/linux/richacl.h @@ -31,7 +31,7 @@ struct richace { }; struct richacl { - atomic_t a_refcount; + struct base_acl a_base; /* must be first, see richacl_put() */ unsigned int a_owner_mask; unsigned int a_group_mask; unsigned int a_other_mask; @@ -56,8 +56,7 @@ struct richacl { static inline struct richacl * richacl_get(struct richacl *acl) { - if (acl) - atomic_inc(&acl->a_refcount); + base_acl_get(&acl->a_base); return acl; } @@ -67,8 +66,8 @@ richacl_get(struct richacl *acl) static inline void richacl_put(struct richacl *acl) { - if (acl && atomic_dec_and_test(&acl->a_refcount)) - kfree(acl); + BUILD_BUG_ON(offsetof(struct richacl, a_base) != 0); + base_acl_put(&acl->a_base); } /**