From patchwork Thu Apr 19 13:33:51 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 10350183 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 3E3AA6023A for ; Thu, 19 Apr 2018 13:36:13 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 2CB7824603 for ; Thu, 19 Apr 2018 13:36:13 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 214CB26E40; Thu, 19 Apr 2018 13:36:13 +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=-7.9 required=2.0 tests=BAYES_00, MAILING_LIST_MULTI, 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 20CE324603 for ; Thu, 19 Apr 2018 13:36:12 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752427AbeDSNd5 (ORCPT ); Thu, 19 Apr 2018 09:33:57 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:43558 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752322AbeDSNdx (ORCPT ); Thu, 19 Apr 2018 09:33:53 -0400 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.rdu2.redhat.com [10.11.54.5]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id E835440201A6; Thu, 19 Apr 2018 13:33:52 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-60.rdu2.redhat.com [10.10.121.60]) by smtp.corp.redhat.com (Postfix) with ESMTP id D5BE01C723; Thu, 19 Apr 2018 13:33:51 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [PATCH 23/24] afs: Implement namespacing [ver #7] From: David Howells To: viro@zeniv.linux.org.uk Cc: linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org, dhowells@redhat.com, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-afs@lists.infradead.org Date: Thu, 19 Apr 2018 14:33:51 +0100 Message-ID: <152414483136.23902.14446781798566249159.stgit@warthog.procyon.org.uk> In-Reply-To: <152414466005.23902.12967974041384198114.stgit@warthog.procyon.org.uk> References: <152414466005.23902.12967974041384198114.stgit@warthog.procyon.org.uk> User-Agent: StGit/0.17.1-dirty MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.11.54.5 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.11.55.6]); Thu, 19 Apr 2018 13:33:52 +0000 (UTC) X-Greylist: inspected by milter-greylist-4.5.16 (mx1.redhat.com [10.11.55.6]); Thu, 19 Apr 2018 13:33:52 +0000 (UTC) for IP:'10.11.54.5' DOMAIN:'int-mx05.intmail.prod.int.rdu2.redhat.com' HELO:'smtp.corp.redhat.com' FROM:'dhowells@redhat.com' RCPT:'' Sender: linux-nfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-nfs@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP Implement namespacing within AFS, but don't yet let mounts occur outside the init namespace. An additional patch will be required propagate the network namespace across automounts. Signed-off-by: David Howells --- fs/afs/cell.c | 4 +- fs/afs/internal.h | 36 ++++++++++++--------- fs/afs/main.c | 33 ++++++++++++++++---- fs/afs/proc.c | 89 +++++++++++++++++++++++++++++++++++------------------ fs/afs/super.c | 58 +++++++++++++++++++++++++---------- 5 files changed, 149 insertions(+), 71 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/fs/afs/cell.c b/fs/afs/cell.c index fdf4c36cff79..a98a8a3d5544 100644 --- a/fs/afs/cell.c +++ b/fs/afs/cell.c @@ -528,7 +528,7 @@ static int afs_activate_cell(struct afs_net *net, struct afs_cell *cell) NULL, 0, cell, 0, true); #endif - ret = afs_proc_cell_setup(net, cell); + ret = afs_proc_cell_setup(cell); if (ret < 0) return ret; spin_lock(&net->proc_cells_lock); @@ -544,7 +544,7 @@ static void afs_deactivate_cell(struct afs_net *net, struct afs_cell *cell) { _enter("%s", cell->name); - afs_proc_cell_remove(net, cell); + afs_proc_cell_remove(cell); spin_lock(&net->proc_cells_lock); list_del_init(&cell->proc_link); diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 0266730b3ad7..a5161c0ae3ab 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -22,6 +22,8 @@ #include #include #include +#include +#include #include #include "afs.h" @@ -192,7 +194,7 @@ struct afs_read { * - there's one superblock per volume */ struct afs_super_info { - struct afs_net *net; /* Network namespace */ + struct net *net_ns; /* Network namespace */ struct afs_cell *cell; /* The cell in which the volume resides */ struct afs_volume *volume; /* volume record */ bool dyn_root; /* True if dynamic root */ @@ -221,6 +223,7 @@ struct afs_sysnames { * AFS network namespace record. */ struct afs_net { + struct net *net; /* Backpointer to the owning net namespace */ struct afs_uuid uuid; bool live; /* F if this namespace is being removed */ @@ -283,7 +286,6 @@ struct afs_net { }; extern const char afs_init_sysname[]; -extern struct afs_net __afs_net;// Dummy AFS network namespace; TODO: replace with real netns enum afs_cell_state { AFS_CELL_UNSET, @@ -790,34 +792,36 @@ extern int afs_drop_inode(struct inode *); * main.c */ extern struct workqueue_struct *afs_wq; +extern int afs_net_id; -static inline struct afs_net *afs_d2net(struct dentry *dentry) +static inline struct afs_net *afs_net(struct net *net) { - return &__afs_net; + return net_generic(net, afs_net_id); } -static inline struct afs_net *afs_i2net(struct inode *inode) +static inline struct afs_net *afs_sb2net(struct super_block *sb) { - return &__afs_net; + return afs_net(AFS_FS_S(sb)->net_ns); } -static inline struct afs_net *afs_v2net(struct afs_vnode *vnode) +static inline struct afs_net *afs_d2net(struct dentry *dentry) { - return &__afs_net; + return afs_sb2net(dentry->d_sb); } -static inline struct afs_net *afs_sock2net(struct sock *sk) +static inline struct afs_net *afs_i2net(struct inode *inode) { - return &__afs_net; + return afs_sb2net(inode->i_sb); } -static inline struct afs_net *afs_get_net(struct afs_net *net) +static inline struct afs_net *afs_v2net(struct afs_vnode *vnode) { - return net; + return afs_i2net(&vnode->vfs_inode); } -static inline void afs_put_net(struct afs_net *net) +static inline struct afs_net *afs_sock2net(struct sock *sk) { + return net_generic(sock_net(sk), afs_net_id); } static inline void __afs_stat(atomic_t *s) @@ -852,8 +856,8 @@ extern int afs_get_ipv4_interfaces(struct afs_interface *, size_t, bool); */ extern int __net_init afs_proc_init(struct afs_net *); extern void __net_exit afs_proc_cleanup(struct afs_net *); -extern int afs_proc_cell_setup(struct afs_net *, struct afs_cell *); -extern void afs_proc_cell_remove(struct afs_net *, struct afs_cell *); +extern int afs_proc_cell_setup(struct afs_cell *); +extern void afs_proc_cell_remove(struct afs_cell *); extern void afs_put_sysnames(struct afs_sysnames *); /* @@ -986,7 +990,7 @@ extern bool afs_annotate_server_list(struct afs_server_list *, struct afs_server * super.c */ extern int __init afs_fs_init(void); -extern void __exit afs_fs_exit(void); +extern void afs_fs_exit(void); /* * vlclient.c diff --git a/fs/afs/main.c b/fs/afs/main.c index d7560168b3bf..7d2c1354e2ca 100644 --- a/fs/afs/main.c +++ b/fs/afs/main.c @@ -15,6 +15,7 @@ #include #include #include +#include #define CREATE_TRACE_POINTS #include "internal.h" @@ -32,7 +33,7 @@ module_param(rootcell, charp, 0); MODULE_PARM_DESC(rootcell, "root AFS cell name and VL server IP addr list"); struct workqueue_struct *afs_wq; -struct afs_net __afs_net; +static struct proc_dir_entry *afs_proc_symlink; #if defined(CONFIG_ALPHA) const char afs_init_sysname[] = "alpha_linux26"; @@ -67,11 +68,13 @@ const char afs_init_sysname[] = "unknown_linux26"; /* * Initialise an AFS network namespace record. */ -static int __net_init afs_net_init(struct afs_net *net) +static int __net_init afs_net_init(struct net *net_ns) { struct afs_sysnames *sysnames; + struct afs_net *net = afs_net(net_ns); int ret; + net->net = net_ns; net->live = true; generate_random_uuid((unsigned char *)&net->uuid); @@ -142,8 +145,10 @@ static int __net_init afs_net_init(struct afs_net *net) /* * Clean up and destroy an AFS network namespace record. */ -static void __net_exit afs_net_exit(struct afs_net *net) +static void __net_exit afs_net_exit(struct net *net_ns) { + struct afs_net *net = afs_net(net_ns); + net->live = false; afs_cell_purge(net); afs_purge_servers(net); @@ -152,6 +157,13 @@ static void __net_exit afs_net_exit(struct afs_net *net) afs_put_sysnames(net->sysnames); } +static struct pernet_operations afs_net_ops = { + .init = afs_net_init, + .exit = afs_net_exit, + .id = &afs_net_id, + .size = sizeof(struct afs_net), +}; + /* * initialise the AFS client FS module */ @@ -178,7 +190,7 @@ static int __init afs_init(void) goto error_cache; #endif - ret = afs_net_init(&__afs_net); + ret = register_pernet_subsys(&afs_net_ops); if (ret < 0) goto error_net; @@ -187,10 +199,18 @@ static int __init afs_init(void) if (ret < 0) goto error_fs; + afs_proc_symlink = proc_symlink("fs/afs", NULL, "../self/net/afs"); + if (IS_ERR(afs_proc_symlink)) { + ret = PTR_ERR(afs_proc_symlink); + goto error_proc; + } + return ret; +error_proc: + afs_fs_exit(); error_fs: - afs_net_exit(&__afs_net); + unregister_pernet_subsys(&afs_net_ops); error_net: #ifdef CONFIG_AFS_FSCACHE fscache_unregister_netfs(&afs_cache_netfs); @@ -219,8 +239,9 @@ static void __exit afs_exit(void) { printk(KERN_INFO "kAFS: Red Hat AFS client v0.1 unregistering.\n"); + proc_remove(afs_proc_symlink); afs_fs_exit(); - afs_net_exit(&__afs_net); + unregister_pernet_subsys(&afs_net_ops); #ifdef CONFIG_AFS_FSCACHE fscache_unregister_netfs(&afs_cache_netfs); #endif diff --git a/fs/afs/proc.c b/fs/afs/proc.c index 839a22280606..cc7c48a5b743 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -17,14 +17,16 @@ #include #include "internal.h" -static inline struct afs_net *afs_proc2net(struct file *f) +static inline struct afs_net *afs_proc2net_get(struct file *f) { - return &__afs_net; + struct net *net_ns = get_proc_net(file_inode(f)); + + return net_ns ? afs_net(net_ns) : NULL; } static inline struct afs_net *afs_seq2net(struct seq_file *m) { - return &__afs_net; // TODO: use seq_file_net(m) + return afs_net(seq_file_net(m)); } static int afs_proc_cells_open(struct inode *inode, struct file *file); @@ -161,7 +163,7 @@ int afs_proc_init(struct afs_net *net) { _enter(""); - net->proc_afs = proc_mkdir("fs/afs", NULL); + net->proc_afs = proc_net_mkdir(net->net, "afs", net->net->proc_net); if (!net->proc_afs) goto error_dir; @@ -196,16 +198,8 @@ void afs_proc_cleanup(struct afs_net *net) */ static int afs_proc_cells_open(struct inode *inode, struct file *file) { - struct seq_file *m; - int ret; - - ret = seq_open(file, &afs_proc_cells_ops); - if (ret < 0) - return ret; - - m = file->private_data; - m->private = PDE_DATA(inode); - return 0; + return seq_open_net(inode, file, &afs_proc_cells_ops, + sizeof(struct seq_net_private)); } /* @@ -266,7 +260,8 @@ static int afs_proc_cells_show(struct seq_file *m, void *v) static ssize_t afs_proc_cells_write(struct file *file, const char __user *buf, size_t size, loff_t *_pos) { - struct afs_net *net = afs_proc2net(file); + struct afs_net *net; + struct net *net_ns = NULL; char *kbuf, *name, *args; int ret; @@ -305,6 +300,12 @@ static ssize_t afs_proc_cells_write(struct file *file, const char __user *buf, /* determine command to perform */ _debug("cmd=%s name=%s args=%s", kbuf, name, args); + ret = -ESTALE; + net_ns = get_proc_net(file_inode(file)); + if (!net_ns) + goto done; + net = afs_net(net_ns); + if (strcmp(kbuf, "add") == 0) { struct afs_cell *cell; @@ -324,6 +325,7 @@ static ssize_t afs_proc_cells_write(struct file *file, const char __user *buf, ret = size; done: + put_net(net_ns); kfree(kbuf); _leave(" = %d", ret); return ret; @@ -338,15 +340,24 @@ static ssize_t afs_proc_rootcell_read(struct file *file, char __user *buf, size_t size, loff_t *_pos) { struct afs_cell *cell; - struct afs_net *net = afs_proc2net(file); + struct afs_net *net; + struct net *net_ns = NULL; unsigned int seq = 0; char name[AFS_MAXCELLNAME + 1]; int len; if (*_pos > 0) return 0; - if (!net->ws_cell) - return 0; + + net_ns = get_proc_net(file_inode(file)); + if (!net_ns) + return -ESTALE; + net = afs_net(net_ns); + + if (!net->ws_cell) { + len = 0; + goto out; + } rcu_read_lock(); do { @@ -362,14 +373,18 @@ static ssize_t afs_proc_rootcell_read(struct file *file, char __user *buf, rcu_read_unlock(); if (!len) - return 0; + goto out; name[len++] = '\n'; if (len > size) len = size; - if (copy_to_user(buf, name, len) != 0) - return -EFAULT; + if (copy_to_user(buf, name, len) != 0) { + len = -EFAULT; + goto out; + } *_pos = 1; +out: + put_net(net_ns); return len; } @@ -381,7 +396,8 @@ static ssize_t afs_proc_rootcell_write(struct file *file, const char __user *buf, size_t size, loff_t *_pos) { - struct afs_net *net = afs_proc2net(file); + struct afs_net *net; + struct net *net_ns = NULL; char *kbuf, *s; int ret; @@ -407,6 +423,12 @@ static ssize_t afs_proc_rootcell_write(struct file *file, /* determine command to perform */ _debug("rootcell=%s", kbuf); + ret = -ESTALE; + net_ns = get_proc_net(file_inode(file)); + if (!net_ns) + goto out; + net = afs_net(net_ns); + ret = afs_cell_init(net, kbuf); if (ret >= 0) ret = size; /* consume everything, always */ @@ -420,13 +442,14 @@ static ssize_t afs_proc_rootcell_write(struct file *file, /* * initialise /proc/fs/afs// */ -int afs_proc_cell_setup(struct afs_net *net, struct afs_cell *cell) +int afs_proc_cell_setup(struct afs_cell *cell) { struct proc_dir_entry *dir; + struct afs_net *net = cell->net; _enter("%p{%s},%p", cell, cell->name, net->proc_afs); - dir = proc_mkdir(cell->name, net->proc_afs); + dir = proc_net_mkdir(net->net, cell->name, net->proc_afs); if (!dir) goto error_dir; @@ -449,12 +472,12 @@ int afs_proc_cell_setup(struct afs_net *net, struct afs_cell *cell) /* * remove /proc/fs/afs// */ -void afs_proc_cell_remove(struct afs_net *net, struct afs_cell *cell) +void afs_proc_cell_remove(struct afs_cell *cell) { - _enter(""); + struct afs_net *net = cell->net; + _enter(""); remove_proc_subtree(cell->name, net->proc_afs); - _leave(""); } @@ -471,7 +494,8 @@ static int afs_proc_cell_volumes_open(struct inode *inode, struct file *file) if (!cell) return -ENOENT; - ret = seq_open(file, &afs_proc_cell_volumes_ops); + ret = seq_open_net(inode, file, &afs_proc_cell_volumes_ops, + sizeof(struct seq_net_private)); if (ret < 0) return ret; @@ -560,7 +584,8 @@ static int afs_proc_cell_vlservers_open(struct inode *inode, struct file *file) if (!cell) return -ENOENT; - ret = seq_open(file, &afs_proc_cell_vlservers_ops); + ret = seq_open_net(inode, file, &afs_proc_cell_vlservers_ops, + sizeof(struct seq_net_private)); if (ret<0) return ret; @@ -649,7 +674,8 @@ static int afs_proc_cell_vlservers_show(struct seq_file *m, void *v) */ static int afs_proc_servers_open(struct inode *inode, struct file *file) { - return seq_open(file, &afs_proc_servers_ops); + return seq_open_net(inode, file, &afs_proc_servers_ops, + sizeof(struct seq_net_private)); } /* @@ -729,7 +755,8 @@ static int afs_proc_sysname_open(struct inode *inode, struct file *file) struct seq_file *m; int ret; - ret = seq_open(file, &afs_proc_sysname_ops); + ret = seq_open_net(inode, file, &afs_proc_sysname_ops, + sizeof(struct seq_net_private)); if (ret < 0) return ret; diff --git a/fs/afs/super.c b/fs/afs/super.c index 6ab0b79e061e..f56070a9c606 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -48,6 +48,8 @@ struct file_system_type afs_fs_type = { }; MODULE_ALIAS_FS("afs"); +int afs_net_id; + static const struct super_operations afs_super_ops = { .statfs = afs_statfs, .alloc_inode = afs_alloc_inode, @@ -117,7 +119,7 @@ int __init afs_fs_init(void) /* * clean up the filesystem */ -void __exit afs_fs_exit(void) +void afs_fs_exit(void) { _enter(""); @@ -391,7 +393,7 @@ static int afs_test_super(struct super_block *sb, struct fs_context *fc) struct afs_fs_context *ctx = container_of(fc, struct afs_fs_context, fc); struct afs_super_info *as = AFS_FS_S(sb); - return (as->net == ctx->net && + return (as->net_ns == ctx->fc.net_ns && as->volume && as->volume->vid == ctx->volume->vid); } @@ -477,7 +479,7 @@ static struct afs_super_info *afs_alloc_sbi(struct afs_fs_context *ctx) as = kzalloc(sizeof(struct afs_super_info), GFP_KERNEL); if (as) { - as->net = afs_get_net(ctx->net); + as->net_ns = get_net(ctx->fc.net_ns); if (ctx->dyn_root) { as->dyn_root = true; } else { @@ -492,8 +494,8 @@ static void afs_destroy_sbi(struct afs_super_info *as) { if (as) { afs_put_volume(as->cell, as->volume); - afs_put_cell(as->net, as->cell); - afs_put_net(as->net); + afs_put_cell(afs_net(as->net_ns), as->cell); + put_net(as->net_ns); kfree(as); } } @@ -506,7 +508,8 @@ static void afs_kill_super(struct super_block *sb) * deactivating the superblock. */ if (as->volume) - afs_clear_callback_interests(as->net, as->volume->servers); + afs_clear_callback_interests(afs_net(as->net_ns), + as->volume->servers); kill_anon_super(sb); if (as->volume) afs_deactivate_volume(as->volume); @@ -574,7 +577,6 @@ static void afs_free_fc(struct fs_context *fc) afs_put_volume(ctx->cell, ctx->volume); afs_put_cell(ctx->net, ctx->cell); - afs_put_net(ctx->net); afs_destroy_sbi(ctx->as); key_put(ctx->key); } @@ -595,19 +597,19 @@ static int afs_init_fs_context(struct fs_context *fc, struct super_block *src_sb struct afs_fs_context *ctx = container_of(fc, struct afs_fs_context, fc); struct afs_super_info *src_as; struct afs_cell *cell; + struct net *net_ns; if (current->nsproxy->net_ns != &init_net) return -EINVAL; + ctx->type = AFSVL_ROVOL; - if (src_sb) { - src_as = AFS_FS_S(src_sb); - if (src_as) { - ctx->net = afs_get_net(src_as->net); - ctx->cell = afs_get_cell(src_as->cell); - ctx->volume = __afs_get_volume(src_as->volume); - } - } else { - ctx->net = afs_get_net(&__afs_net); + switch (ctx->fc.purpose) { + case FS_CONTEXT_FOR_USER_MOUNT: + case FS_CONTEXT_FOR_KERNEL_MOUNT: + ctx->fc.net_ns = maybe_get_net(current->nsproxy->net_ns); + if (!ctx->fc.net_ns) + return -ESTALE; + ctx->net = afs_net(ctx->fc.net_ns); /* Default to the workstation cell. */ rcu_read_lock(); @@ -616,6 +618,30 @@ static int afs_init_fs_context(struct fs_context *fc, struct super_block *src_sb if (IS_ERR(cell)) cell = NULL; ctx->cell = cell; + break; + + case FS_CONTEXT_FOR_SUBMOUNT: + if (!src_sb) + return -EINVAL; + + src_as = AFS_FS_S(src_sb); + ASSERT(src_as); + + net_ns = maybe_get_net(src_as->net_ns); + if (!net_ns) + return -ESTALE; + ctx->fc.net_ns = net_ns; + ctx->net = afs_net(net_ns); + if (src_as->cell) + ctx->cell = afs_get_cell(src_as->cell); + if (src_as->volume && src_as->volume->type == AFSVL_RWVOL) { + ctx->type = AFSVL_RWVOL; + ctx->force = true; + } + break; + + case FS_CONTEXT_FOR_RECONFIGURE: + break; } ctx->fc.ops = &afs_context_ops;