From 1729a16c2c92bbd9e54ac7cad3101fea2e073aa5 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Wed, 26 Nov 2008 12:03:54 +0100 Subject: fuse: style fixes Fix coding style errors reported by checkpatch and others. Uptdate copyright date to 2008. Signed-off-by: Miklos Szeredi --- fs/fuse/control.c | 6 ++++-- fs/fuse/dev.c | 22 +++++++++++++--------- fs/fuse/dir.c | 12 +++++++----- fs/fuse/file.c | 4 ++-- fs/fuse/fuse_i.h | 38 +++++++++++++++++++------------------- fs/fuse/inode.c | 19 +++++++++---------- 6 files changed, 54 insertions(+), 47 deletions(-) (limited to 'fs') diff --git a/fs/fuse/control.c b/fs/fuse/control.c index 4f3cab32141..99c99dfb037 100644 --- a/fs/fuse/control.c +++ b/fs/fuse/control.c @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2006 Miklos Szeredi + Copyright (C) 2001-2008 Miklos Szeredi This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -48,11 +48,13 @@ static ssize_t fuse_conn_waiting_read(struct file *file, char __user *buf, size_t size; if (!*ppos) { + long value; struct fuse_conn *fc = fuse_ctl_file_conn_get(file); if (!fc) return 0; - file->private_data=(void *)(long)atomic_read(&fc->num_waiting); + value = atomic_read(&fc->num_waiting); + file->private_data = (void *)value; fuse_conn_put(fc); } size = sprintf(tmp, "%ld\n", (long)file->private_data); diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index b72361479be..85a23bb524f 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2006 Miklos Szeredi + Copyright (C) 2001-2008 Miklos Szeredi This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -539,8 +539,8 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) BUG_ON(!cs->nr_segs); cs->seglen = cs->iov[0].iov_len; cs->addr = (unsigned long) cs->iov[0].iov_base; - cs->iov ++; - cs->nr_segs --; + cs->iov++; + cs->nr_segs--; } down_read(¤t->mm->mmap_sem); err = get_user_pages(current, current->mm, cs->addr, 1, cs->write, 0, @@ -589,9 +589,11 @@ static int fuse_copy_page(struct fuse_copy_state *cs, struct page *page, kunmap_atomic(mapaddr, KM_USER1); } while (count) { - int err; - if (!cs->len && (err = fuse_copy_fill(cs))) - return err; + if (!cs->len) { + int err = fuse_copy_fill(cs); + if (err) + return err; + } if (page) { void *mapaddr = kmap_atomic(page, KM_USER1); void *buf = mapaddr + offset; @@ -631,9 +633,11 @@ static int fuse_copy_pages(struct fuse_copy_state *cs, unsigned nbytes, static int fuse_copy_one(struct fuse_copy_state *cs, void *val, unsigned size) { while (size) { - int err; - if (!cs->len && (err = fuse_copy_fill(cs))) - return err; + if (!cs->len) { + int err = fuse_copy_fill(cs); + if (err) + return err; + } fuse_copy_do(cs, &val, &size); } return 0; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index fd03330cade..9e7c5385699 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2006 Miklos Szeredi + Copyright (C) 2001-2008 Miklos Szeredi This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -204,7 +204,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd) return 0; } spin_lock(&fc->lock); - fi->nlookup ++; + fi->nlookup++; spin_unlock(&fc->lock); } fuse_put_request(fc, forget_req); @@ -637,9 +637,11 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry) if (!err) { struct inode *inode = entry->d_inode; - /* Set nlink to zero so the inode can be cleared, if - the inode does have more links this will be - discovered at the next lookup/getattr */ + /* + * Set nlink to zero so the inode can be cleared, if the inode + * does have more links this will be discovered at the next + * lookup/getattr. + */ clear_nlink(inode); fuse_invalidate_attr(inode); fuse_invalidate_attr(dir); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 34930a964b8..86054f437d1 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2006 Miklos Szeredi + Copyright (C) 2001-2008 Miklos Szeredi This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -543,7 +543,7 @@ static int fuse_readpages_fill(void *_data, struct page *page) } } req->pages[req->num_pages] = page; - req->num_pages ++; + req->num_pages++; return 0; } diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 35accfdd747..4fc5131f5c9 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2006 Miklos Szeredi + Copyright (C) 2001-2008 Miklos Szeredi This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -355,19 +355,19 @@ struct fuse_conn { /** Connection failed (version mismatch). Cannot race with setting other bitfields since it is only set once in INIT reply, before any other request, and never cleared */ - unsigned conn_error : 1; + unsigned conn_error:1; /** Connection successful. Only set in INIT */ - unsigned conn_init : 1; + unsigned conn_init:1; /** Do readpages asynchronously? Only set in INIT */ - unsigned async_read : 1; + unsigned async_read:1; /** Do not send separate SETATTR request before open(O_TRUNC) */ - unsigned atomic_o_trunc : 1; + unsigned atomic_o_trunc:1; /** Filesystem supports NFS exporting. Only set in INIT */ - unsigned export_support : 1; + unsigned export_support:1; /* * The following bitfields are only for optimization purposes @@ -375,43 +375,43 @@ struct fuse_conn { */ /** Is fsync not implemented by fs? */ - unsigned no_fsync : 1; + unsigned no_fsync:1; /** Is fsyncdir not implemented by fs? */ - unsigned no_fsyncdir : 1; + unsigned no_fsyncdir:1; /** Is flush not implemented by fs? */ - unsigned no_flush : 1; + unsigned no_flush:1; /** Is setxattr not implemented by fs? */ - unsigned no_setxattr : 1; + unsigned no_setxattr:1; /** Is getxattr not implemented by fs? */ - unsigned no_getxattr : 1; + unsigned no_getxattr:1; /** Is listxattr not implemented by fs? */ - unsigned no_listxattr : 1; + unsigned no_listxattr:1; /** Is removexattr not implemented by fs? */ - unsigned no_removexattr : 1; + unsigned no_removexattr:1; /** Are file locking primitives not implemented by fs? */ - unsigned no_lock : 1; + unsigned no_lock:1; /** Is access not implemented by fs? */ - unsigned no_access : 1; + unsigned no_access:1; /** Is create not implemented by fs? */ - unsigned no_create : 1; + unsigned no_create:1; /** Is interrupt not implemented by fs? */ - unsigned no_interrupt : 1; + unsigned no_interrupt:1; /** Is bmap not implemented by fs? */ - unsigned no_bmap : 1; + unsigned no_bmap:1; /** Do multi-page cached writes */ - unsigned big_writes : 1; + unsigned big_writes:1; /** The number of requests waiting for completion */ atomic_t num_waiting; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 2e99f34b443..739595b4196 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2006 Miklos Szeredi + Copyright (C) 2001-2008 Miklos Szeredi This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -37,10 +37,10 @@ struct fuse_mount_data { unsigned rootmode; unsigned user_id; unsigned group_id; - unsigned fd_present : 1; - unsigned rootmode_present : 1; - unsigned user_id_present : 1; - unsigned group_id_present : 1; + unsigned fd_present:1; + unsigned rootmode_present:1; + unsigned user_id_present:1; + unsigned group_id_present:1; unsigned flags; unsigned max_read; unsigned blksize; @@ -250,7 +250,7 @@ struct inode *fuse_iget(struct super_block *sb, u64 nodeid, fi = get_fuse_inode(inode); spin_lock(&fc->lock); - fi->nlookup ++; + fi->nlookup++; spin_unlock(&fc->lock); fuse_change_attributes(inode, attr, attr_valid, attr_version); @@ -553,8 +553,7 @@ static struct inode *get_root_inode(struct super_block *sb, unsigned mode) return fuse_iget(sb, 1, 0, &attr, 0, 0); } -struct fuse_inode_handle -{ +struct fuse_inode_handle { u64 nodeid; u32 generation; }; @@ -952,7 +951,7 @@ static inline void unregister_fuseblk(void) static void fuse_inode_init_once(void *foo) { - struct inode * inode = foo; + struct inode *inode = foo; inode_init_once(inode); } @@ -1031,7 +1030,7 @@ static int __init fuse_init(void) { int res; - printk("fuse init (API version %i.%i)\n", + printk(KERN_INFO "fuse init (API version %i.%i)\n", FUSE_KERNEL_VERSION, FUSE_KERNEL_MINOR_VERSION); INIT_LIST_HEAD(&fuse_conn_list); -- cgit v1.2.3-70-g09d2 From e9bb09dd6c5b8ec6a971ed6251df5eba3a4c8d3c Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:54 +0100 Subject: fuse: don't let fuse_req->end() put the base reference fuse_req->end() was supposed to be put the base reference but there's no reason why it should. It only makes things more complex. Move it out of ->end() and make it the responsibility of request_end(). Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 5 ++--- fs/fuse/file.c | 5 ++--- fs/fuse/inode.c | 1 - 3 files changed, 4 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 85a23bb524f..225388f54ae 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -293,8 +293,7 @@ static void request_end(struct fuse_conn *fc, struct fuse_req *req) wake_up(&req->waitq); if (end) end(fc, req); - else - fuse_put_request(fc, req); + fuse_put_request(fc, req); } static void wait_answer_interruptible(struct fuse_conn *fc, @@ -1006,11 +1005,11 @@ static void end_io_requests(struct fuse_conn *fc) wake_up(&req->waitq); if (end) { req->end = NULL; - /* The end function will consume this reference */ __fuse_get_request(req); spin_unlock(&fc->lock); wait_event(req->waitq, !req->locked); end(fc, req); + fuse_put_request(fc, req); spin_lock(&fc->lock); } } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 86054f437d1..61726980391 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -79,7 +79,6 @@ static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req) { dput(req->misc.release.dentry); mntput(req->misc.release.vfsmount); - fuse_put_request(fc, req); } static void fuse_file_put(struct fuse_file *ff) @@ -493,7 +492,6 @@ static void fuse_readpages_end(struct fuse_conn *fc, struct fuse_req *req) } if (req->ff) fuse_file_put(req->ff); - fuse_put_request(fc, req); } static void fuse_send_readpages(struct fuse_req *req, struct file *file, @@ -513,6 +511,7 @@ static void fuse_send_readpages(struct fuse_req *req, struct file *file, } else { request_send(fc, req); fuse_readpages_end(fc, req); + fuse_put_request(fc, req); } } @@ -1042,7 +1041,6 @@ static void fuse_writepage_free(struct fuse_conn *fc, struct fuse_req *req) { __free_page(req->pages[0]); fuse_file_put(req->ff); - fuse_put_request(fc, req); } static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req) @@ -1086,6 +1084,7 @@ static void fuse_send_writepage(struct fuse_conn *fc, struct fuse_req *req) fuse_writepage_finish(fc, req); spin_unlock(&fc->lock); fuse_writepage_free(fc, req); + fuse_put_request(fc, req); spin_lock(&fc->lock); } diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 739595b4196..fa474989ec7 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -760,7 +760,6 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) fc->max_write = max_t(unsigned, 4096, fc->max_write); fc->conn_init = 1; } - fuse_put_request(fc, req); fc->blocked = 0; wake_up_all(&fc->blocked_waitq); } -- cgit v1.2.3-70-g09d2 From 59efec7b903987dcb60b9ebc85c7acd4443a11a1 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:55 +0100 Subject: fuse: implement ioctl support Generic ioctl support is tricky to implement because only the ioctl implementation itself knows which memory regions need to be read and/or written. To support this, fuse client can request retry of ioctl specifying memory regions to read and write. Deep copying (nested pointers) can be implemented by retrying multiple times resolving one depth of dereference at a time. For security and cleanliness considerations, ioctl implementation has restricted mode where the kernel determines data transfer directions and sizes using the _IOC_*() macros on the ioctl command. In this mode, retry is not allowed. For all FUSE servers, restricted mode is enforced. Unrestricted ioctl will be used by CUSE. Plese read the comment on top of fs/fuse/file.c::fuse_file_do_ioctl() for more information. Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 280 +++++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/fuse.h | 32 ++++++ 2 files changed, 312 insertions(+) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 61726980391..baed06ea762 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1469,6 +1469,282 @@ static loff_t fuse_file_llseek(struct file *file, loff_t offset, int origin) return retval; } +static int fuse_ioctl_copy_user(struct page **pages, struct iovec *iov, + unsigned int nr_segs, size_t bytes, bool to_user) +{ + struct iov_iter ii; + int page_idx = 0; + + if (!bytes) + return 0; + + iov_iter_init(&ii, iov, nr_segs, bytes, 0); + + while (iov_iter_count(&ii)) { + struct page *page = pages[page_idx++]; + size_t todo = min_t(size_t, PAGE_SIZE, iov_iter_count(&ii)); + void *kaddr, *map; + + kaddr = map = kmap(page); + + while (todo) { + char __user *uaddr = ii.iov->iov_base + ii.iov_offset; + size_t iov_len = ii.iov->iov_len - ii.iov_offset; + size_t copy = min(todo, iov_len); + size_t left; + + if (!to_user) + left = copy_from_user(kaddr, uaddr, copy); + else + left = copy_to_user(uaddr, kaddr, copy); + + if (unlikely(left)) + return -EFAULT; + + iov_iter_advance(&ii, copy); + todo -= copy; + kaddr += copy; + } + + kunmap(map); + } + + return 0; +} + +/* + * For ioctls, there is no generic way to determine how much memory + * needs to be read and/or written. Furthermore, ioctls are allowed + * to dereference the passed pointer, so the parameter requires deep + * copying but FUSE has no idea whatsoever about what to copy in or + * out. + * + * This is solved by allowing FUSE server to retry ioctl with + * necessary in/out iovecs. Let's assume the ioctl implementation + * needs to read in the following structure. + * + * struct a { + * char *buf; + * size_t buflen; + * } + * + * On the first callout to FUSE server, inarg->in_size and + * inarg->out_size will be NULL; then, the server completes the ioctl + * with FUSE_IOCTL_RETRY set in out->flags, out->in_iovs set to 1 and + * the actual iov array to + * + * { { .iov_base = inarg.arg, .iov_len = sizeof(struct a) } } + * + * which tells FUSE to copy in the requested area and retry the ioctl. + * On the second round, the server has access to the structure and + * from that it can tell what to look for next, so on the invocation, + * it sets FUSE_IOCTL_RETRY, out->in_iovs to 2 and iov array to + * + * { { .iov_base = inarg.arg, .iov_len = sizeof(struct a) }, + * { .iov_base = a.buf, .iov_len = a.buflen } } + * + * FUSE will copy both struct a and the pointed buffer from the + * process doing the ioctl and retry ioctl with both struct a and the + * buffer. + * + * This time, FUSE server has everything it needs and completes ioctl + * without FUSE_IOCTL_RETRY which finishes the ioctl call. + * + * Copying data out works the same way. + * + * Note that if FUSE_IOCTL_UNRESTRICTED is clear, the kernel + * automatically initializes in and out iovs by decoding @cmd with + * _IOC_* macros and the server is not allowed to request RETRY. This + * limits ioctl data transfers to well-formed ioctls and is the forced + * behavior for all FUSE servers. + */ +static long fuse_file_do_ioctl(struct file *file, unsigned int cmd, + unsigned long arg, unsigned int flags) +{ + struct inode *inode = file->f_dentry->d_inode; + struct fuse_file *ff = file->private_data; + struct fuse_conn *fc = get_fuse_conn(inode); + struct fuse_ioctl_in inarg = { + .fh = ff->fh, + .cmd = cmd, + .arg = arg, + .flags = flags + }; + struct fuse_ioctl_out outarg; + struct fuse_req *req = NULL; + struct page **pages = NULL; + struct page *iov_page = NULL; + struct iovec *in_iov = NULL, *out_iov = NULL; + unsigned int in_iovs = 0, out_iovs = 0, num_pages = 0, max_pages; + size_t in_size, out_size, transferred; + int err; + + /* assume all the iovs returned by client always fits in a page */ + BUILD_BUG_ON(sizeof(struct iovec) * FUSE_IOCTL_MAX_IOV > PAGE_SIZE); + + if (!fuse_allow_task(fc, current)) + return -EACCES; + + err = -EIO; + if (is_bad_inode(inode)) + goto out; + + err = -ENOMEM; + pages = kzalloc(sizeof(pages[0]) * FUSE_MAX_PAGES_PER_REQ, GFP_KERNEL); + iov_page = alloc_page(GFP_KERNEL); + if (!pages || !iov_page) + goto out; + + /* + * If restricted, initialize IO parameters as encoded in @cmd. + * RETRY from server is not allowed. + */ + if (!(flags & FUSE_IOCTL_UNRESTRICTED)) { + struct iovec *iov = page_address(iov_page); + + iov->iov_base = (void *)arg; + iov->iov_len = _IOC_SIZE(cmd); + + if (_IOC_DIR(cmd) & _IOC_WRITE) { + in_iov = iov; + in_iovs = 1; + } + + if (_IOC_DIR(cmd) & _IOC_READ) { + out_iov = iov; + out_iovs = 1; + } + } + + retry: + inarg.in_size = in_size = iov_length(in_iov, in_iovs); + inarg.out_size = out_size = iov_length(out_iov, out_iovs); + + /* + * Out data can be used either for actual out data or iovs, + * make sure there always is at least one page. + */ + out_size = max_t(size_t, out_size, PAGE_SIZE); + max_pages = DIV_ROUND_UP(max(in_size, out_size), PAGE_SIZE); + + /* make sure there are enough buffer pages and init request with them */ + err = -ENOMEM; + if (max_pages > FUSE_MAX_PAGES_PER_REQ) + goto out; + while (num_pages < max_pages) { + pages[num_pages] = alloc_page(GFP_KERNEL | __GFP_HIGHMEM); + if (!pages[num_pages]) + goto out; + num_pages++; + } + + req = fuse_get_req(fc); + if (IS_ERR(req)) { + err = PTR_ERR(req); + req = NULL; + goto out; + } + memcpy(req->pages, pages, sizeof(req->pages[0]) * num_pages); + req->num_pages = num_pages; + + /* okay, let's send it to the client */ + req->in.h.opcode = FUSE_IOCTL; + req->in.h.nodeid = get_node_id(inode); + req->in.numargs = 1; + req->in.args[0].size = sizeof(inarg); + req->in.args[0].value = &inarg; + if (in_size) { + req->in.numargs++; + req->in.args[1].size = in_size; + req->in.argpages = 1; + + err = fuse_ioctl_copy_user(pages, in_iov, in_iovs, in_size, + false); + if (err) + goto out; + } + + req->out.numargs = 2; + req->out.args[0].size = sizeof(outarg); + req->out.args[0].value = &outarg; + req->out.args[1].size = out_size; + req->out.argpages = 1; + req->out.argvar = 1; + + request_send(fc, req); + err = req->out.h.error; + transferred = req->out.args[1].size; + fuse_put_request(fc, req); + req = NULL; + if (err) + goto out; + + /* did it ask for retry? */ + if (outarg.flags & FUSE_IOCTL_RETRY) { + char *vaddr; + + /* no retry if in restricted mode */ + err = -EIO; + if (!(flags & FUSE_IOCTL_UNRESTRICTED)) + goto out; + + in_iovs = outarg.in_iovs; + out_iovs = outarg.out_iovs; + + /* + * Make sure things are in boundary, separate checks + * are to protect against overflow. + */ + err = -ENOMEM; + if (in_iovs > FUSE_IOCTL_MAX_IOV || + out_iovs > FUSE_IOCTL_MAX_IOV || + in_iovs + out_iovs > FUSE_IOCTL_MAX_IOV) + goto out; + + err = -EIO; + if ((in_iovs + out_iovs) * sizeof(struct iovec) != transferred) + goto out; + + /* okay, copy in iovs and retry */ + vaddr = kmap_atomic(pages[0], KM_USER0); + memcpy(page_address(iov_page), vaddr, transferred); + kunmap_atomic(vaddr, KM_USER0); + + in_iov = page_address(iov_page); + out_iov = in_iov + in_iovs; + + goto retry; + } + + err = -EIO; + if (transferred > inarg.out_size) + goto out; + + err = fuse_ioctl_copy_user(pages, out_iov, out_iovs, transferred, true); + out: + if (req) + fuse_put_request(fc, req); + if (iov_page) + __free_page(iov_page); + while (num_pages) + __free_page(pages[--num_pages]); + kfree(pages); + + return err ? err : outarg.result; +} + +static long fuse_file_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + return fuse_file_do_ioctl(file, cmd, arg, 0); +} + +static long fuse_file_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + return fuse_file_do_ioctl(file, cmd, arg, FUSE_IOCTL_COMPAT); +} + static const struct file_operations fuse_file_operations = { .llseek = fuse_file_llseek, .read = do_sync_read, @@ -1483,6 +1759,8 @@ static const struct file_operations fuse_file_operations = { .lock = fuse_file_lock, .flock = fuse_file_flock, .splice_read = generic_file_splice_read, + .unlocked_ioctl = fuse_file_ioctl, + .compat_ioctl = fuse_file_compat_ioctl, }; static const struct file_operations fuse_direct_io_file_operations = { @@ -1495,6 +1773,8 @@ static const struct file_operations fuse_direct_io_file_operations = { .fsync = fuse_fsync, .lock = fuse_file_lock, .flock = fuse_file_flock, + .unlocked_ioctl = fuse_file_ioctl, + .compat_ioctl = fuse_file_compat_ioctl, /* no mmap and splice_read */ }; diff --git a/include/linux/fuse.h b/include/linux/fuse.h index 7caa473306e..608e300ae88 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -148,6 +148,21 @@ struct fuse_file_lock { */ #define FUSE_READ_LOCKOWNER (1 << 1) +/** + * Ioctl flags + * + * FUSE_IOCTL_COMPAT: 32bit compat ioctl on 64bit machine + * FUSE_IOCTL_UNRESTRICTED: not restricted to well-formed ioctls, retry allowed + * FUSE_IOCTL_RETRY: retry with new iovecs + * + * FUSE_IOCTL_MAX_IOV: maximum of in_iovecs + out_iovecs + */ +#define FUSE_IOCTL_COMPAT (1 << 0) +#define FUSE_IOCTL_UNRESTRICTED (1 << 1) +#define FUSE_IOCTL_RETRY (1 << 2) + +#define FUSE_IOCTL_MAX_IOV 256 + enum fuse_opcode { FUSE_LOOKUP = 1, FUSE_FORGET = 2, /* no reply */ @@ -185,6 +200,7 @@ enum fuse_opcode { FUSE_INTERRUPT = 36, FUSE_BMAP = 37, FUSE_DESTROY = 38, + FUSE_IOCTL = 39, }; /* The read buffer is required to be at least 8k, but may be much larger */ @@ -385,6 +401,22 @@ struct fuse_bmap_out { __u64 block; }; +struct fuse_ioctl_in { + __u64 fh; + __u32 flags; + __u32 cmd; + __u64 arg; + __u32 in_size; + __u32 out_size; +}; + +struct fuse_ioctl_out { + __s32 result; + __u32 flags; + __u32 in_iovs; + __u32 out_iovs; +}; + struct fuse_in_header { __u32 len; __u32 opcode; -- cgit v1.2.3-70-g09d2 From acf99433d98c2570a619d8fb8b51abce4e532059 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:55 +0100 Subject: fuse: add file kernel handle The file handle, fuse_file->fh, is opaque value supplied by userland FUSE server and uniqueness is not guaranteed. Add file kernel handle, fuse_file->kh, which is allocated by the kernel on file allocation and guaranteed to be unique. This will be used by poll to match notification to the respective file but can be used for other purposes where unique file handle is necessary. Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 2 +- fs/fuse/file.c | 8 ++++++-- fs/fuse/fuse_i.h | 8 +++++++- fs/fuse/inode.c | 1 + 4 files changed, 15 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 9e7c5385699..16ae55d347b 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -408,7 +408,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, goto out_put_forget_req; err = -ENOMEM; - ff = fuse_file_alloc(); + ff = fuse_file_alloc(fc); if (!ff) goto out_put_request; diff --git a/fs/fuse/file.c b/fs/fuse/file.c index baed06ea762..a28ced678d3 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -46,7 +46,7 @@ static int fuse_send_open(struct inode *inode, struct file *file, int isdir, return err; } -struct fuse_file *fuse_file_alloc(void) +struct fuse_file *fuse_file_alloc(struct fuse_conn *fc) { struct fuse_file *ff; ff = kmalloc(sizeof(struct fuse_file), GFP_KERNEL); @@ -58,6 +58,9 @@ struct fuse_file *fuse_file_alloc(void) } else { INIT_LIST_HEAD(&ff->write_entry); atomic_set(&ff->count, 0); + spin_lock(&fc->lock); + ff->kh = ++fc->khctr; + spin_unlock(&fc->lock); } } return ff; @@ -108,6 +111,7 @@ void fuse_finish_open(struct inode *inode, struct file *file, int fuse_open_common(struct inode *inode, struct file *file, int isdir) { + struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_open_out outarg; struct fuse_file *ff; int err; @@ -120,7 +124,7 @@ int fuse_open_common(struct inode *inode, struct file *file, int isdir) if (err) return err; - ff = fuse_file_alloc(); + ff = fuse_file_alloc(fc); if (!ff) return -ENOMEM; diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 4fc5131f5c9..86f01330382 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -100,6 +100,9 @@ struct fuse_file { /** Request reserved for flush and release */ struct fuse_req *reserved_req; + /** Kernel file handle guaranteed to be unique */ + u64 kh; + /** File handle used by userspace */ u64 fh; @@ -322,6 +325,9 @@ struct fuse_conn { /** The list of requests under I/O */ struct list_head io; + /** The next unique kernel file handle */ + u64 khctr; + /** Number of requests currently in the background */ unsigned num_background; @@ -499,7 +505,7 @@ void fuse_read_fill(struct fuse_req *req, struct file *file, */ int fuse_open_common(struct inode *inode, struct file *file, int isdir); -struct fuse_file *fuse_file_alloc(void); +struct fuse_file *fuse_file_alloc(struct fuse_conn *fc); void fuse_file_free(struct fuse_file *ff); void fuse_finish_open(struct inode *inode, struct file *file, struct fuse_file *ff, struct fuse_open_out *outarg); diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index fa474989ec7..0e15bc221d2 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -485,6 +485,7 @@ static struct fuse_conn *new_conn(struct super_block *sb) fc->bdi.unplug_io_fn = default_unplug_io_fn; /* fuse does it's own writeback accounting */ fc->bdi.capabilities = BDI_CAP_NO_ACCT_WB; + fc->khctr = 0; fc->dev = sb->s_dev; err = bdi_init(&fc->bdi); if (err) -- cgit v1.2.3-70-g09d2 From 8599396b5062bf6bd2a0b433503849e2322df1c2 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:55 +0100 Subject: fuse: implement unsolicited notification Clients always used to write only in response to read requests. To implement poll efficiently, clients should be able to issue unsolicited notifications. This patch implements basic notification support. Zero fuse_out_header.unique is now accepted and considered unsolicited notification and the error field contains notification code. This patch doesn't implement any actual notification. Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 27 +++++++++++++++++++++++++-- include/linux/fuse.h | 4 ++++ 2 files changed, 29 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 225388f54ae..ffd670bb8c8 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -816,6 +816,15 @@ static ssize_t fuse_dev_read(struct kiocb *iocb, const struct iovec *iov, return err; } +static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, + unsigned int size, struct fuse_copy_state *cs) +{ + switch (code) { + default: + return -EINVAL; + } +} + /* Look up request on processing list by unique ID */ static struct fuse_req *request_find(struct fuse_conn *fc, u64 unique) { @@ -879,9 +888,23 @@ static ssize_t fuse_dev_write(struct kiocb *iocb, const struct iovec *iov, err = fuse_copy_one(&cs, &oh, sizeof(oh)); if (err) goto err_finish; + + err = -EINVAL; + if (oh.len != nbytes) + goto err_finish; + + /* + * Zero oh.unique indicates unsolicited notification message + * and error contains notification code. + */ + if (!oh.unique) { + err = fuse_notify(fc, oh.error, nbytes - sizeof(oh), &cs); + fuse_copy_finish(&cs); + return err ? err : nbytes; + } + err = -EINVAL; - if (!oh.unique || oh.error <= -1000 || oh.error > 0 || - oh.len != nbytes) + if (oh.error <= -1000 || oh.error > 0) goto err_finish; spin_lock(&fc->lock); diff --git a/include/linux/fuse.h b/include/linux/fuse.h index 608e300ae88..abde9949e2c 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -203,6 +203,10 @@ enum fuse_opcode { FUSE_IOCTL = 39, }; +enum fuse_notify_code { + FUSE_NOTIFY_CODE_MAX, +}; + /* The read buffer is required to be at least 8k, but may be much larger */ #define FUSE_MIN_READ_BUFFER 8192 -- cgit v1.2.3-70-g09d2 From 95668a69a4bb862063c4d28a746e55107dee7b98 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:55 +0100 Subject: fuse: implement poll support Implement poll support. Polled files are indexed using kh in a RB tree rooted at fuse_conn->polled_files. Client should send FUSE_NOTIFY_POLL notification once after processing FUSE_POLL which has FUSE_POLL_SCHEDULE_NOTIFY set. Sending notification unconditionally after the latest poll or everytime file content might have changed is inefficient but won't cause malfunction. fuse_file_poll() can sleep and requires patches from the following thread which allows f_op->poll() to sleep. http://thread.gmane.org/gmane.linux.kernel/726176 Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 19 ++++++++ fs/fuse/file.c | 132 +++++++++++++++++++++++++++++++++++++++++++++++++++ fs/fuse/fuse_i.h | 20 ++++++++ fs/fuse/inode.c | 1 + include/linux/fuse.h | 25 ++++++++++ 5 files changed, 197 insertions(+) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index ffd670bb8c8..6176e444c76 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -816,10 +816,29 @@ static ssize_t fuse_dev_read(struct kiocb *iocb, const struct iovec *iov, return err; } +static int fuse_notify_poll(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_poll_wakeup_out outarg; + int err; + + if (size != sizeof(outarg)) + return -EINVAL; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + return err; + + return fuse_notify_poll_wakeup(fc, &outarg); +} + static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, unsigned int size, struct fuse_copy_state *cs) { switch (code) { + case FUSE_NOTIFY_POLL: + return fuse_notify_poll(fc, size, cs); + default: return -EINVAL; } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index a28ced678d3..b3a944e4bb9 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -62,6 +62,8 @@ struct fuse_file *fuse_file_alloc(struct fuse_conn *fc) ff->kh = ++fc->khctr; spin_unlock(&fc->lock); } + RB_CLEAR_NODE(&ff->polled_node); + init_waitqueue_head(&ff->poll_wait); } return ff; } @@ -170,7 +172,11 @@ int fuse_release_common(struct inode *inode, struct file *file, int isdir) spin_lock(&fc->lock); list_del(&ff->write_entry); + if (!RB_EMPTY_NODE(&ff->polled_node)) + rb_erase(&ff->polled_node, &fc->polled_files); spin_unlock(&fc->lock); + + wake_up_interruptible_sync(&ff->poll_wait); /* * Normally this will send the RELEASE request, * however if some asynchronous READ or WRITE requests @@ -1749,6 +1755,130 @@ static long fuse_file_compat_ioctl(struct file *file, unsigned int cmd, return fuse_file_do_ioctl(file, cmd, arg, FUSE_IOCTL_COMPAT); } +/* + * All files which have been polled are linked to RB tree + * fuse_conn->polled_files which is indexed by kh. Walk the tree and + * find the matching one. + */ +static struct rb_node **fuse_find_polled_node(struct fuse_conn *fc, u64 kh, + struct rb_node **parent_out) +{ + struct rb_node **link = &fc->polled_files.rb_node; + struct rb_node *last = NULL; + + while (*link) { + struct fuse_file *ff; + + last = *link; + ff = rb_entry(last, struct fuse_file, polled_node); + + if (kh < ff->kh) + link = &last->rb_left; + else if (kh > ff->kh) + link = &last->rb_right; + else + return link; + } + + if (parent_out) + *parent_out = last; + return link; +} + +/* + * The file is about to be polled. Make sure it's on the polled_files + * RB tree. Note that files once added to the polled_files tree are + * not removed before the file is released. This is because a file + * polled once is likely to be polled again. + */ +static void fuse_register_polled_file(struct fuse_conn *fc, + struct fuse_file *ff) +{ + spin_lock(&fc->lock); + if (RB_EMPTY_NODE(&ff->polled_node)) { + struct rb_node **link, *parent; + + link = fuse_find_polled_node(fc, ff->kh, &parent); + BUG_ON(*link); + rb_link_node(&ff->polled_node, parent, link); + rb_insert_color(&ff->polled_node, &fc->polled_files); + } + spin_unlock(&fc->lock); +} + +static unsigned fuse_file_poll(struct file *file, poll_table *wait) +{ + struct inode *inode = file->f_dentry->d_inode; + struct fuse_file *ff = file->private_data; + struct fuse_conn *fc = get_fuse_conn(inode); + struct fuse_poll_in inarg = { .fh = ff->fh, .kh = ff->kh }; + struct fuse_poll_out outarg; + struct fuse_req *req; + int err; + + if (fc->no_poll) + return DEFAULT_POLLMASK; + + poll_wait(file, &ff->poll_wait, wait); + + /* + * Ask for notification iff there's someone waiting for it. + * The client may ignore the flag and always notify. + */ + if (waitqueue_active(&ff->poll_wait)) { + inarg.flags |= FUSE_POLL_SCHEDULE_NOTIFY; + fuse_register_polled_file(fc, ff); + } + + req = fuse_get_req(fc); + if (IS_ERR(req)) + return PTR_ERR(req); + + req->in.h.opcode = FUSE_POLL; + req->in.h.nodeid = get_node_id(inode); + req->in.numargs = 1; + req->in.args[0].size = sizeof(inarg); + req->in.args[0].value = &inarg; + req->out.numargs = 1; + req->out.args[0].size = sizeof(outarg); + req->out.args[0].value = &outarg; + request_send(fc, req); + err = req->out.h.error; + fuse_put_request(fc, req); + + if (!err) + return outarg.revents; + if (err == -ENOSYS) { + fc->no_poll = 1; + return DEFAULT_POLLMASK; + } + return POLLERR; +} + +/* + * This is called from fuse_handle_notify() on FUSE_NOTIFY_POLL and + * wakes up the poll waiters. + */ +int fuse_notify_poll_wakeup(struct fuse_conn *fc, + struct fuse_notify_poll_wakeup_out *outarg) +{ + u64 kh = outarg->kh; + struct rb_node **link; + + spin_lock(&fc->lock); + + link = fuse_find_polled_node(fc, kh, NULL); + if (*link) { + struct fuse_file *ff; + + ff = rb_entry(*link, struct fuse_file, polled_node); + wake_up_interruptible_sync(&ff->poll_wait); + } + + spin_unlock(&fc->lock); + return 0; +} + static const struct file_operations fuse_file_operations = { .llseek = fuse_file_llseek, .read = do_sync_read, @@ -1765,6 +1895,7 @@ static const struct file_operations fuse_file_operations = { .splice_read = generic_file_splice_read, .unlocked_ioctl = fuse_file_ioctl, .compat_ioctl = fuse_file_compat_ioctl, + .poll = fuse_file_poll, }; static const struct file_operations fuse_direct_io_file_operations = { @@ -1779,6 +1910,7 @@ static const struct file_operations fuse_direct_io_file_operations = { .flock = fuse_file_flock, .unlocked_ioctl = fuse_file_ioctl, .compat_ioctl = fuse_file_compat_ioctl, + .poll = fuse_file_poll, /* no mmap and splice_read */ }; diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 86f01330382..986fbd4c1ff 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -19,6 +19,8 @@ #include #include #include +#include +#include /** Max number of pages that can be used in a single read request */ #define FUSE_MAX_PAGES_PER_REQ 32 @@ -111,6 +113,12 @@ struct fuse_file { /** Entry on inode's write_files list */ struct list_head write_entry; + + /** RB node to be linked on fuse_conn->polled_files */ + struct rb_node polled_node; + + /** Wait queue head for poll */ + wait_queue_head_t poll_wait; }; /** One input argument of a request */ @@ -328,6 +336,9 @@ struct fuse_conn { /** The next unique kernel file handle */ u64 khctr; + /** rbtree of fuse_files waiting for poll events indexed by ph */ + struct rb_root polled_files; + /** Number of requests currently in the background */ unsigned num_background; @@ -416,6 +427,9 @@ struct fuse_conn { /** Is bmap not implemented by fs? */ unsigned no_bmap:1; + /** Is poll not implemented by fs? */ + unsigned no_poll:1; + /** Do multi-page cached writes */ unsigned big_writes:1; @@ -524,6 +538,12 @@ int fuse_release_common(struct inode *inode, struct file *file, int isdir); int fuse_fsync_common(struct file *file, struct dentry *de, int datasync, int isdir); +/** + * Notify poll wakeup + */ +int fuse_notify_poll_wakeup(struct fuse_conn *fc, + struct fuse_notify_poll_wakeup_out *outarg); + /** * Initialize file operations on a regular file */ diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 0e15bc221d2..ba725612808 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -486,6 +486,7 @@ static struct fuse_conn *new_conn(struct super_block *sb) /* fuse does it's own writeback accounting */ fc->bdi.capabilities = BDI_CAP_NO_ACCT_WB; fc->khctr = 0; + fc->polled_files = RB_ROOT; fc->dev = sb->s_dev; err = bdi_init(&fc->bdi); if (err) diff --git a/include/linux/fuse.h b/include/linux/fuse.h index abde9949e2c..5650cf033e7 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -163,6 +163,13 @@ struct fuse_file_lock { #define FUSE_IOCTL_MAX_IOV 256 +/** + * Poll flags + * + * FUSE_POLL_SCHEDULE_NOTIFY: request poll notify + */ +#define FUSE_POLL_SCHEDULE_NOTIFY (1 << 0) + enum fuse_opcode { FUSE_LOOKUP = 1, FUSE_FORGET = 2, /* no reply */ @@ -201,9 +208,11 @@ enum fuse_opcode { FUSE_BMAP = 37, FUSE_DESTROY = 38, FUSE_IOCTL = 39, + FUSE_POLL = 40, }; enum fuse_notify_code { + FUSE_NOTIFY_POLL = 1, FUSE_NOTIFY_CODE_MAX, }; @@ -421,6 +430,22 @@ struct fuse_ioctl_out { __u32 out_iovs; }; +struct fuse_poll_in { + __u64 fh; + __u64 kh; + __u32 flags; + __u32 padding; +}; + +struct fuse_poll_out { + __u32 revents; + __u32 padding; +}; + +struct fuse_notify_poll_wakeup_out { + __u64 kh; +}; + struct fuse_in_header { __u32 len; __u32 opcode; -- cgit v1.2.3-70-g09d2 From b93f858ab2a4bee779c360002f313ad6c3504cdc Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:55 +0100 Subject: fuse: add fuse_ prefix to several functions Add fuse_ prefix to request_send*() and get_root_inode() as some of those functions will be exported for CUSE. With or without CUSE export, having the function names scoped is a good idea for debuggability. Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 23 ++++++++++++----------- fs/fuse/dir.c | 34 +++++++++++++++++----------------- fs/fuse/file.c | 28 ++++++++++++++-------------- fs/fuse/fuse_i.h | 9 +++++---- fs/fuse/inode.c | 12 ++++++------ 5 files changed, 54 insertions(+), 52 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 6176e444c76..3c44ce359a0 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -379,7 +379,7 @@ static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) } } -void request_send(struct fuse_conn *fc, struct fuse_req *req) +void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) { req->isreply = 1; spin_lock(&fc->lock); @@ -398,8 +398,8 @@ void request_send(struct fuse_conn *fc, struct fuse_req *req) spin_unlock(&fc->lock); } -static void request_send_nowait_locked(struct fuse_conn *fc, - struct fuse_req *req) +static void fuse_request_send_nowait_locked(struct fuse_conn *fc, + struct fuse_req *req) { req->background = 1; fc->num_background++; @@ -413,11 +413,11 @@ static void request_send_nowait_locked(struct fuse_conn *fc, flush_bg_queue(fc); } -static void request_send_nowait(struct fuse_conn *fc, struct fuse_req *req) +static void fuse_request_send_nowait(struct fuse_conn *fc, struct fuse_req *req) { spin_lock(&fc->lock); if (fc->connected) { - request_send_nowait_locked(fc, req); + fuse_request_send_nowait_locked(fc, req); spin_unlock(&fc->lock); } else { req->out.h.error = -ENOTCONN; @@ -425,16 +425,16 @@ static void request_send_nowait(struct fuse_conn *fc, struct fuse_req *req) } } -void request_send_noreply(struct fuse_conn *fc, struct fuse_req *req) +void fuse_request_send_noreply(struct fuse_conn *fc, struct fuse_req *req) { req->isreply = 0; - request_send_nowait(fc, req); + fuse_request_send_nowait(fc, req); } -void request_send_background(struct fuse_conn *fc, struct fuse_req *req) +void fuse_request_send_background(struct fuse_conn *fc, struct fuse_req *req) { req->isreply = 1; - request_send_nowait(fc, req); + fuse_request_send_nowait(fc, req); } /* @@ -442,10 +442,11 @@ void request_send_background(struct fuse_conn *fc, struct fuse_req *req) * * fc->connected must have been checked previously */ -void request_send_background_locked(struct fuse_conn *fc, struct fuse_req *req) +void fuse_request_send_background_locked(struct fuse_conn *fc, + struct fuse_req *req) { req->isreply = 1; - request_send_nowait_locked(fc, req); + fuse_request_send_nowait_locked(fc, req); } /* diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 16ae55d347b..f310768a02f 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -189,7 +189,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd) parent = dget_parent(entry); fuse_lookup_init(fc, req, get_node_id(parent->d_inode), &entry->d_name, &outarg); - request_send(fc, req); + fuse_request_send(fc, req); dput(parent); err = req->out.h.error; fuse_put_request(fc, req); @@ -283,7 +283,7 @@ int fuse_lookup_name(struct super_block *sb, u64 nodeid, struct qstr *name, attr_version = fuse_get_attr_version(fc); fuse_lookup_init(fc, req, nodeid, name, outarg); - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); /* Zero nodeid is same as -ENOENT, but with valid timeout */ @@ -369,7 +369,7 @@ static void fuse_sync_release(struct fuse_conn *fc, struct fuse_file *ff, { fuse_release_fill(ff, nodeid, flags, FUSE_RELEASE); ff->reserved_req->force = 1; - request_send(fc, ff->reserved_req); + fuse_request_send(fc, ff->reserved_req); fuse_put_request(fc, ff->reserved_req); kfree(ff); } @@ -432,7 +432,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, req->out.args[0].value = &outentry; req->out.args[1].size = sizeof(outopen); req->out.args[1].value = &outopen; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; if (err) { if (err == -ENOSYS) @@ -502,7 +502,7 @@ static int create_new_entry(struct fuse_conn *fc, struct fuse_req *req, else req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err) @@ -631,7 +631,7 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry) req->in.numargs = 1; req->in.args[0].size = entry->d_name.len + 1; req->in.args[0].value = entry->d_name.name; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (!err) { @@ -664,7 +664,7 @@ static int fuse_rmdir(struct inode *dir, struct dentry *entry) req->in.numargs = 1; req->in.args[0].size = entry->d_name.len + 1; req->in.args[0].value = entry->d_name.name; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (!err) { @@ -697,7 +697,7 @@ static int fuse_rename(struct inode *olddir, struct dentry *oldent, req->in.args[1].value = oldent->d_name.name; req->in.args[2].size = newent->d_name.len + 1; req->in.args[2].value = newent->d_name.name; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (!err) { @@ -813,7 +813,7 @@ static int fuse_do_getattr(struct inode *inode, struct kstat *stat, else req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (!err) { @@ -906,7 +906,7 @@ static int fuse_access(struct inode *inode, int mask) req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err == -ENOSYS) { @@ -1028,7 +1028,7 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) req->num_pages = 1; req->pages[0] = page; fuse_read_fill(req, file, inode, file->f_pos, PAGE_SIZE, FUSE_READDIR); - request_send(fc, req); + fuse_request_send(fc, req); nbytes = req->out.args[0].size; err = req->out.h.error; fuse_put_request(fc, req); @@ -1062,7 +1062,7 @@ static char *read_link(struct dentry *dentry) req->out.numargs = 1; req->out.args[0].size = PAGE_SIZE - 1; req->out.args[0].value = link; - request_send(fc, req); + fuse_request_send(fc, req); if (req->out.h.error) { free_page((unsigned long) link); link = ERR_PTR(req->out.h.error); @@ -1268,7 +1268,7 @@ static int fuse_do_setattr(struct dentry *entry, struct iattr *attr, else req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err) { @@ -1362,7 +1362,7 @@ static int fuse_setxattr(struct dentry *entry, const char *name, req->in.args[1].value = name; req->in.args[2].size = size; req->in.args[2].value = value; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err == -ENOSYS) { @@ -1408,7 +1408,7 @@ static ssize_t fuse_getxattr(struct dentry *entry, const char *name, req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; } - request_send(fc, req); + fuse_request_send(fc, req); ret = req->out.h.error; if (!ret) ret = size ? req->out.args[0].size : outarg.size; @@ -1458,7 +1458,7 @@ static ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size) req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; } - request_send(fc, req); + fuse_request_send(fc, req); ret = req->out.h.error; if (!ret) ret = size ? req->out.args[0].size : outarg.size; @@ -1491,7 +1491,7 @@ static int fuse_removexattr(struct dentry *entry, const char *name) req->in.numargs = 1; req->in.args[0].size = strlen(name) + 1; req->in.args[0].value = name; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err == -ENOSYS) { diff --git a/fs/fuse/file.c b/fs/fuse/file.c index b3a944e4bb9..80b5fa80f5e 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -39,7 +39,7 @@ static int fuse_send_open(struct inode *inode, struct file *file, int isdir, req->out.numargs = 1; req->out.args[0].size = sizeof(*outargp); req->out.args[0].value = outargp; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); @@ -93,7 +93,7 @@ static void fuse_file_put(struct fuse_file *ff) struct inode *inode = req->misc.release.dentry->d_inode; struct fuse_conn *fc = get_fuse_conn(inode); req->end = fuse_release_end; - request_send_background(fc, req); + fuse_request_send_background(fc, req); kfree(ff); } } @@ -289,7 +289,7 @@ static int fuse_flush(struct file *file, fl_owner_t id) req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; req->force = 1; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err == -ENOSYS) { @@ -353,7 +353,7 @@ int fuse_fsync_common(struct file *file, struct dentry *de, int datasync, req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err == -ENOSYS) { @@ -405,7 +405,7 @@ static size_t fuse_send_read(struct fuse_req *req, struct file *file, inarg->read_flags |= FUSE_READ_LOCKOWNER; inarg->lock_owner = fuse_lock_owner_id(fc, owner); } - request_send(fc, req); + fuse_request_send(fc, req); return req->out.args[0].size; } @@ -517,9 +517,9 @@ static void fuse_send_readpages(struct fuse_req *req, struct file *file, struct fuse_file *ff = file->private_data; req->ff = fuse_file_get(ff); req->end = fuse_readpages_end; - request_send_background(fc, req); + fuse_request_send_background(fc, req); } else { - request_send(fc, req); + fuse_request_send(fc, req); fuse_readpages_end(fc, req); fuse_put_request(fc, req); } @@ -645,7 +645,7 @@ static size_t fuse_send_write(struct fuse_req *req, struct file *file, inarg->write_flags |= FUSE_WRITE_LOCKOWNER; inarg->lock_owner = fuse_lock_owner_id(fc, owner); } - request_send(fc, req); + fuse_request_send(fc, req); return req->misc.write.out.size; } @@ -1087,7 +1087,7 @@ static void fuse_send_writepage(struct fuse_conn *fc, struct fuse_req *req) req->in.args[1].size = inarg->size; fi->writectr++; - request_send_background_locked(fc, req); + fuse_request_send_background_locked(fc, req); return; out_free: @@ -1334,7 +1334,7 @@ static int fuse_getlk(struct file *file, struct file_lock *fl) req->out.numargs = 1; req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (!err) @@ -1366,7 +1366,7 @@ static int fuse_setlk(struct file *file, struct file_lock *fl, int flock) return PTR_ERR(req); fuse_lk_fill(req, file, fl, opcode, pid, flock); - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; /* locking is restartable */ if (err == -EINTR) @@ -1442,7 +1442,7 @@ static sector_t fuse_bmap(struct address_space *mapping, sector_t block) req->out.numargs = 1; req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); if (err == -ENOSYS) @@ -1681,7 +1681,7 @@ static long fuse_file_do_ioctl(struct file *file, unsigned int cmd, req->out.argpages = 1; req->out.argvar = 1; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; transferred = req->out.args[1].size; fuse_put_request(fc, req); @@ -1842,7 +1842,7 @@ static unsigned fuse_file_poll(struct file *file, poll_table *wait) req->out.numargs = 1; req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 986fbd4c1ff..c08e7e89092 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -619,19 +619,20 @@ void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req); /** * Send a request (synchronous) */ -void request_send(struct fuse_conn *fc, struct fuse_req *req); +void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req); /** * Send a request with no reply */ -void request_send_noreply(struct fuse_conn *fc, struct fuse_req *req); +void fuse_request_send_noreply(struct fuse_conn *fc, struct fuse_req *req); /** * Send a request in the background */ -void request_send_background(struct fuse_conn *fc, struct fuse_req *req); +void fuse_request_send_background(struct fuse_conn *fc, struct fuse_req *req); -void request_send_background_locked(struct fuse_conn *fc, struct fuse_req *req); +void fuse_request_send_background_locked(struct fuse_conn *fc, + struct fuse_req *req); /* Abort all requests */ void fuse_abort_conn(struct fuse_conn *fc); diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index ba725612808..ee91639c6d3 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -94,7 +94,7 @@ void fuse_send_forget(struct fuse_conn *fc, struct fuse_req *req, req->in.numargs = 1; req->in.args[0].size = sizeof(struct fuse_forget_in); req->in.args[0].value = inarg; - request_send_noreply(fc, req); + fuse_request_send_noreply(fc, req); } static void fuse_clear_inode(struct inode *inode) @@ -269,7 +269,7 @@ static void fuse_send_destroy(struct fuse_conn *fc) fc->destroy_req = NULL; req->in.h.opcode = FUSE_DESTROY; req->force = 1; - request_send(fc, req); + fuse_request_send(fc, req); fuse_put_request(fc, req); } } @@ -334,7 +334,7 @@ static int fuse_statfs(struct dentry *dentry, struct kstatfs *buf) req->out.args[0].size = fc->minor < 4 ? FUSE_COMPAT_STATFS_SIZE : sizeof(outarg); req->out.args[0].value = &outarg; - request_send(fc, req); + fuse_request_send(fc, req); err = req->out.h.error; if (!err) convert_fuse_statfs(buf, &outarg.st); @@ -544,7 +544,7 @@ struct fuse_conn *fuse_conn_get(struct fuse_conn *fc) return fc; } -static struct inode *get_root_inode(struct super_block *sb, unsigned mode) +static struct inode *fuse_get_root_inode(struct super_block *sb, unsigned mode) { struct fuse_attr attr; memset(&attr, 0, sizeof(attr)); @@ -787,7 +787,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) req->out.args[0].size = sizeof(struct fuse_init_out); req->out.args[0].value = &req->misc.init_out; req->end = process_init_reply; - request_send_background(fc, req); + fuse_request_send_background(fc, req); } static int fuse_fill_super(struct super_block *sb, void *data, int silent) @@ -841,7 +841,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) sb->s_fs_info = fc; err = -ENOMEM; - root = get_root_inode(sb, d.rootmode); + root = fuse_get_root_inode(sb, d.rootmode); if (!root) goto err; -- cgit v1.2.3-70-g09d2 From 0d179aa59285ceef529c125e181cbb79ff5245c2 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:55 +0100 Subject: fuse: separate out fuse_conn_init() from new_conn() Separate out fuse_conn_init() from new_conn() and while at it initialize fuse_conn->entry during conn initialization. This will be used by CUSE. Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/fuse_i.h | 5 +++ fs/fuse/inode.c | 119 +++++++++++++++++++++++++++++-------------------------- 2 files changed, 67 insertions(+), 57 deletions(-) (limited to 'fs') diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index c08e7e89092..eb488d48b83 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -649,6 +649,11 @@ void fuse_invalidate_entry_cache(struct dentry *entry); */ struct fuse_conn *fuse_conn_get(struct fuse_conn *fc); +/** + * Initialize fuse_conn + */ +int fuse_conn_init(struct fuse_conn *fc, struct super_block *sb); + /** * Release reference to fuse_conn */ diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index ee91639c6d3..6c9fa03aa36 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -462,70 +462,69 @@ static int fuse_show_options(struct seq_file *m, struct vfsmount *mnt) return 0; } -static struct fuse_conn *new_conn(struct super_block *sb) +int fuse_conn_init(struct fuse_conn *fc, struct super_block *sb) { - struct fuse_conn *fc; int err; - fc = kzalloc(sizeof(*fc), GFP_KERNEL); - if (fc) { - spin_lock_init(&fc->lock); - mutex_init(&fc->inst_mutex); - atomic_set(&fc->count, 1); - init_waitqueue_head(&fc->waitq); - init_waitqueue_head(&fc->blocked_waitq); - init_waitqueue_head(&fc->reserved_req_waitq); - INIT_LIST_HEAD(&fc->pending); - INIT_LIST_HEAD(&fc->processing); - INIT_LIST_HEAD(&fc->io); - INIT_LIST_HEAD(&fc->interrupts); - INIT_LIST_HEAD(&fc->bg_queue); - atomic_set(&fc->num_waiting, 0); - fc->bdi.ra_pages = (VM_MAX_READAHEAD * 1024) / PAGE_CACHE_SIZE; - fc->bdi.unplug_io_fn = default_unplug_io_fn; - /* fuse does it's own writeback accounting */ - fc->bdi.capabilities = BDI_CAP_NO_ACCT_WB; - fc->khctr = 0; - fc->polled_files = RB_ROOT; - fc->dev = sb->s_dev; - err = bdi_init(&fc->bdi); - if (err) - goto error_kfree; - if (sb->s_bdev) { - err = bdi_register(&fc->bdi, NULL, "%u:%u-fuseblk", - MAJOR(fc->dev), MINOR(fc->dev)); - } else { - err = bdi_register_dev(&fc->bdi, fc->dev); - } - if (err) - goto error_bdi_destroy; - /* - * For a single fuse filesystem use max 1% of dirty + - * writeback threshold. - * - * This gives about 1M of write buffer for memory maps on a - * machine with 1G and 10% dirty_ratio, which should be more - * than enough. - * - * Privileged users can raise it by writing to - * - * /sys/class/bdi//max_ratio - */ - bdi_set_max_ratio(&fc->bdi, 1); - fc->reqctr = 0; - fc->blocked = 1; - fc->attr_version = 1; - get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key)); + memset(fc, 0, sizeof(*fc)); + spin_lock_init(&fc->lock); + mutex_init(&fc->inst_mutex); + atomic_set(&fc->count, 1); + init_waitqueue_head(&fc->waitq); + init_waitqueue_head(&fc->blocked_waitq); + init_waitqueue_head(&fc->reserved_req_waitq); + INIT_LIST_HEAD(&fc->pending); + INIT_LIST_HEAD(&fc->processing); + INIT_LIST_HEAD(&fc->io); + INIT_LIST_HEAD(&fc->interrupts); + INIT_LIST_HEAD(&fc->bg_queue); + INIT_LIST_HEAD(&fc->entry); + atomic_set(&fc->num_waiting, 0); + fc->bdi.ra_pages = (VM_MAX_READAHEAD * 1024) / PAGE_CACHE_SIZE; + fc->bdi.unplug_io_fn = default_unplug_io_fn; + /* fuse does it's own writeback accounting */ + fc->bdi.capabilities = BDI_CAP_NO_ACCT_WB; + fc->khctr = 0; + fc->polled_files = RB_ROOT; + fc->dev = sb->s_dev; + err = bdi_init(&fc->bdi); + if (err) + goto error_mutex_destroy; + if (sb->s_bdev) { + err = bdi_register(&fc->bdi, NULL, "%u:%u-fuseblk", + MAJOR(fc->dev), MINOR(fc->dev)); + } else { + err = bdi_register_dev(&fc->bdi, fc->dev); } - return fc; + if (err) + goto error_bdi_destroy; + /* + * For a single fuse filesystem use max 1% of dirty + + * writeback threshold. + * + * This gives about 1M of write buffer for memory maps on a + * machine with 1G and 10% dirty_ratio, which should be more + * than enough. + * + * Privileged users can raise it by writing to + * + * /sys/class/bdi//max_ratio + */ + bdi_set_max_ratio(&fc->bdi, 1); + fc->reqctr = 0; + fc->blocked = 1; + fc->attr_version = 1; + get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key)); + + return 0; -error_bdi_destroy: + error_bdi_destroy: bdi_destroy(&fc->bdi); -error_kfree: + error_mutex_destroy: mutex_destroy(&fc->inst_mutex); - kfree(fc); - return NULL; + return err; } +EXPORT_SYMBOL_GPL(fuse_conn_init); void fuse_conn_put(struct fuse_conn *fc) { @@ -828,10 +827,16 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) if (file->f_op != &fuse_dev_operations) return -EINVAL; - fc = new_conn(sb); + fc = kmalloc(sizeof(*fc), GFP_KERNEL); if (!fc) return -ENOMEM; + err = fuse_conn_init(fc, sb); + if (err) { + kfree(fc); + return err; + } + fc->flags = d.flags; fc->user_id = d.user_id; fc->group_id = d.group_id; -- cgit v1.2.3-70-g09d2 From 43901aabd7a043e62e24e9459dc4949b4cd69f07 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 26 Nov 2008 12:03:56 +0100 Subject: fuse: add fuse_conn->release() Add fuse_conn->release() so that fuse_conn can be embedded in other structures. Signed-off-by: Tejun Heo Signed-off-by: Miklos Szeredi --- fs/fuse/fuse_i.h | 3 +++ fs/fuse/inode.c | 8 +++++++- 2 files changed, 10 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index eb488d48b83..5e64b815a5a 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -465,6 +465,9 @@ struct fuse_conn { /** Version counter for attribute changes */ u64 attr_version; + + /** Called on final put */ + void (*release)(struct fuse_conn *); }; static inline struct fuse_conn *get_fuse_conn_super(struct super_block *sb) diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 6c9fa03aa36..47c96fdca1a 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -533,7 +533,7 @@ void fuse_conn_put(struct fuse_conn *fc) fuse_request_free(fc->destroy_req); mutex_destroy(&fc->inst_mutex); bdi_destroy(&fc->bdi); - kfree(fc); + fc->release(fc); } } @@ -789,6 +789,11 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) fuse_request_send_background(fc, req); } +static void fuse_free_conn(struct fuse_conn *fc) +{ + kfree(fc); +} + static int fuse_fill_super(struct super_block *sb, void *data, int silent) { struct fuse_conn *fc; @@ -837,6 +842,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) return err; } + fc->release = fuse_free_conn; fc->flags = d.flags; fc->user_id = d.user_id; fc->group_id = d.group_id; -- cgit v1.2.3-70-g09d2 From c9f0523d88fd208ce094995a0ac63f7c04e56bab Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 2 Dec 2008 14:49:42 +0100 Subject: fuse: fix sparse warning in ioctl Fix sparse warning: CHECK fs/fuse/file.c fs/fuse/file.c:1615:17: warning: incorrect type in assignment (different address spaces) fs/fuse/file.c:1615:17: expected void [noderef] *iov_base fs/fuse/file.c:1615:17: got void * This was introduced by "fuse: implement ioctl support". Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 80b5fa80f5e..4d2f1339a88 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1612,7 +1612,7 @@ static long fuse_file_do_ioctl(struct file *file, unsigned int cmd, if (!(flags & FUSE_IOCTL_UNRESTRICTED)) { struct iovec *iov = page_address(iov_page); - iov->iov_base = (void *)arg; + iov->iov_base = (void __user *)arg; iov->iov_len = _IOC_SIZE(cmd); if (_IOC_DIR(cmd) & _IOC_WRITE) { -- cgit v1.2.3-70-g09d2 From 5d9ec854bfb6f1e122b1d96b344164a71eac5be8 Mon Sep 17 00:00:00 2001 From: Harvey Harrison Date: Tue, 2 Dec 2008 14:49:42 +0100 Subject: fuse: clean up annotations of fc->lock Makes the existing annotations match the more common one per line style and adds a few missing annotations. Signed-off-by: Harvey Harrison Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 17 ++++++++++++----- fs/fuse/file.c | 4 ++++ 2 files changed, 16 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 3c44ce359a0..69ebf2ecb08 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -269,7 +269,7 @@ static void flush_bg_queue(struct fuse_conn *fc) * Called with fc->lock, unlocks it */ static void request_end(struct fuse_conn *fc, struct fuse_req *req) - __releases(fc->lock) +__releases(&fc->lock) { void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; req->end = NULL; @@ -298,7 +298,8 @@ static void request_end(struct fuse_conn *fc, struct fuse_req *req) static void wait_answer_interruptible(struct fuse_conn *fc, struct fuse_req *req) - __releases(fc->lock) __acquires(fc->lock) +__releases(&fc->lock) +__acquires(&fc->lock) { if (signal_pending(current)) return; @@ -316,7 +317,8 @@ static void queue_interrupt(struct fuse_conn *fc, struct fuse_req *req) } static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) - __releases(fc->lock) __acquires(fc->lock) +__releases(&fc->lock) +__acquires(&fc->lock) { if (!fc->no_interrupt) { /* Any signal may interrupt this */ @@ -668,6 +670,8 @@ static int request_pending(struct fuse_conn *fc) /* Wait until a request is available on the pending list */ static void request_wait(struct fuse_conn *fc) +__releases(&fc->lock) +__acquires(&fc->lock) { DECLARE_WAITQUEUE(wait, current); @@ -695,7 +699,7 @@ static void request_wait(struct fuse_conn *fc) */ static int fuse_read_interrupt(struct fuse_conn *fc, struct fuse_req *req, const struct iovec *iov, unsigned long nr_segs) - __releases(fc->lock) +__releases(&fc->lock) { struct fuse_copy_state cs; struct fuse_in_header ih; @@ -1012,6 +1016,8 @@ static unsigned fuse_dev_poll(struct file *file, poll_table *wait) * This function releases and reacquires fc->lock */ static void end_requests(struct fuse_conn *fc, struct list_head *head) +__releases(&fc->lock) +__acquires(&fc->lock) { while (!list_empty(head)) { struct fuse_req *req; @@ -1034,7 +1040,8 @@ static void end_requests(struct fuse_conn *fc, struct list_head *head) * locked). */ static void end_io_requests(struct fuse_conn *fc) - __releases(fc->lock) __acquires(fc->lock) +__releases(&fc->lock) +__acquires(&fc->lock) { while (!list_empty(&fc->io)) { struct fuse_req *req = diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 4d2f1339a88..1a057f02e7d 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1068,6 +1068,8 @@ static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req) /* Called under fc->lock, may release and reacquire it */ static void fuse_send_writepage(struct fuse_conn *fc, struct fuse_req *req) +__releases(&fc->lock) +__acquires(&fc->lock) { struct fuse_inode *fi = get_fuse_inode(req->inode); loff_t size = i_size_read(req->inode); @@ -1105,6 +1107,8 @@ static void fuse_send_writepage(struct fuse_conn *fc, struct fuse_req *req) * Called with fc->lock */ void fuse_flush_writepages(struct inode *inode) +__releases(&fc->lock) +__acquires(&fc->lock) { struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_inode *fi = get_fuse_inode(inode); -- cgit v1.2.3-70-g09d2 From b4df2b92d8461444fac429c75ba6e125c63056bc Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 27 Oct 2008 22:48:36 +0300 Subject: proc: stop using BKL There are four BKL users in proc: de_put(), proc_lookup_de(), proc_readdir_de(), proc_root_readdir(), 1) de_put() ----------- de_put() is classic atomic_dec_and_test() refcount wrapper -- no BKL needed. BKL doesn't matter to possible refcount leak as well. 2) proc_lookup_de() ------------------- Walking PDE list is protected by proc_subdir_lock(), proc_get_inode() is potentially blocking, all callers of proc_lookup_de() eventually end up from ->lookup hooks which is protected by directory's ->i_mutex -- BKL doesn't protect anything. 3) proc_readdir_de() -------------------- "." and ".." part doesn't need BKL, walking PDE list is under proc_subdir_lock, calling filldir callback is potentially blocking because it writes to luserspace. All proc_readdir_de() callers eventually come from ->readdir hook which is under directory's ->i_mutex -- BKL doesn't protect anything. 4) proc_root_readdir_de() ------------------------- proc_root_readdir_de is ->readdir hook, see (3). Since readdir hooks doesn't use BKL anymore, switch to generic_file_llseek, since it also takes directory's i_mutex. Signed-off-by: Alexey Dobriyan --- fs/proc/generic.c | 8 ++------ fs/proc/inode.c | 3 --- fs/proc/proc_net.c | 2 +- fs/proc/root.c | 8 +------- 4 files changed, 4 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 60a359b3558..db7fa5cab98 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -14,7 +14,6 @@ #include #include #include -#include #include #include #include @@ -379,7 +378,6 @@ struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *dir, struct inode *inode = NULL; int error = -ENOENT; - lock_kernel(); spin_lock(&proc_subdir_lock); for (de = de->subdir; de ; de = de->next) { if (de->namelen != dentry->d_name.len) @@ -397,7 +395,6 @@ struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *dir, } spin_unlock(&proc_subdir_lock); out_unlock: - unlock_kernel(); if (inode) { dentry->d_op = &proc_dentry_operations; @@ -432,8 +429,6 @@ int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent, struct inode *inode = filp->f_path.dentry->d_inode; int ret = 0; - lock_kernel(); - ino = inode->i_ino; i = filp->f_pos; switch (i) { @@ -487,7 +482,7 @@ int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent, spin_unlock(&proc_subdir_lock); } ret = 1; -out: unlock_kernel(); +out: return ret; } @@ -504,6 +499,7 @@ int proc_readdir(struct file *filp, void *dirent, filldir_t filldir) * the /proc directory. */ static const struct file_operations proc_dir_operations = { + .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = proc_readdir, }; diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 2543fd00c65..3e76bb9b3ad 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -35,16 +35,13 @@ struct proc_dir_entry *de_get(struct proc_dir_entry *de) */ void de_put(struct proc_dir_entry *de) { - lock_kernel(); if (!atomic_read(&de->count)) { printk("de_put: entry %s already free!\n", de->name); - unlock_kernel(); return; } if (atomic_dec_and_test(&de->count)) free_proc_entry(de); - unlock_kernel(); } /* diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c index 7bc296f424a..04d1270f1c3 100644 --- a/fs/proc/proc_net.c +++ b/fs/proc/proc_net.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include #include @@ -172,6 +171,7 @@ static int proc_tgid_net_readdir(struct file *filp, void *dirent, } const struct file_operations proc_net_operations = { + .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = proc_tgid_net_readdir, }; diff --git a/fs/proc/root.c b/fs/proc/root.c index 7761602af9d..f6299a25594 100644 --- a/fs/proc/root.c +++ b/fs/proc/root.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include @@ -162,17 +161,12 @@ static int proc_root_readdir(struct file * filp, unsigned int nr = filp->f_pos; int ret; - lock_kernel(); - if (nr < FIRST_PROCESS_ENTRY) { int error = proc_readdir(filp, dirent, filldir); - if (error <= 0) { - unlock_kernel(); + if (error <= 0) return error; - } filp->f_pos = FIRST_PROCESS_ENTRY; } - unlock_kernel(); ret = proc_pid_readdir(filp, dirent, filldir); return ret; -- cgit v1.2.3-70-g09d2 From ecae934edc0c29ec7405da18855004c317de26c6 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sun, 9 Nov 2008 23:12:18 +0300 Subject: proc: remove useless WARN_ONs NULL "struct inode *" means VFS passed NULL inode to ->open. Signed-off-by: Alexey Dobriyan --- fs/proc/base.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index cad92c1ac2b..8642623ea79 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1186,8 +1186,6 @@ static int sched_show(struct seq_file *m, void *v) struct inode *inode = m->private; struct task_struct *p; - WARN_ON(!inode); - p = get_proc_task(inode); if (!p) return -ESRCH; @@ -1205,8 +1203,6 @@ sched_write(struct file *file, const char __user *buf, struct inode *inode = file->f_path.dentry->d_inode; struct task_struct *p; - WARN_ON(!inode); - p = get_proc_task(inode); if (!p) return -ESRCH; -- cgit v1.2.3-70-g09d2 From 631f9c1868b970197747c80fc5168ad7d9fd5d53 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 10 Nov 2008 01:32:52 +0300 Subject: proc: remove '##' usage Inability to jump to /proc/*/foo handlers with ctags is annoying. Signed-off-by: Alexey Dobriyan --- fs/proc/base.c | 183 ++++++++++++++++++++++++++++----------------------------- 1 file changed, 90 insertions(+), 93 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index 8642623ea79..ce7a6da1b6a 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -109,25 +109,22 @@ struct pid_entry { .op = OP, \ } -#define DIR(NAME, MODE, OTYPE) \ - NOD(NAME, (S_IFDIR|(MODE)), \ - &proc_##OTYPE##_inode_operations, &proc_##OTYPE##_operations, \ - {} ) -#define LNK(NAME, OTYPE) \ +#define DIR(NAME, MODE, iops, fops) \ + NOD(NAME, (S_IFDIR|(MODE)), &iops, &fops, {} ) +#define LNK(NAME, get_link) \ NOD(NAME, (S_IFLNK|S_IRWXUGO), \ &proc_pid_link_inode_operations, NULL, \ - { .proc_get_link = &proc_##OTYPE##_link } ) -#define REG(NAME, MODE, OTYPE) \ - NOD(NAME, (S_IFREG|(MODE)), NULL, \ - &proc_##OTYPE##_operations, {}) -#define INF(NAME, MODE, OTYPE) \ + { .proc_get_link = get_link } ) +#define REG(NAME, MODE, fops) \ + NOD(NAME, (S_IFREG|(MODE)), NULL, &fops, {}) +#define INF(NAME, MODE, read) \ NOD(NAME, (S_IFREG|(MODE)), \ NULL, &proc_info_file_operations, \ - { .proc_read = &proc_##OTYPE } ) -#define ONE(NAME, MODE, OTYPE) \ + { .proc_read = read } ) +#define ONE(NAME, MODE, show) \ NOD(NAME, (S_IFREG|(MODE)), \ NULL, &proc_single_file_operations, \ - { .proc_show = &proc_##OTYPE } ) + { .proc_show = show } ) /* * Count the number of hardlinks for the pid_entry table, excluding the . @@ -2134,12 +2131,12 @@ static const struct file_operations proc_pid_attr_operations = { }; static const struct pid_entry attr_dir_stuff[] = { - REG("current", S_IRUGO|S_IWUGO, pid_attr), - REG("prev", S_IRUGO, pid_attr), - REG("exec", S_IRUGO|S_IWUGO, pid_attr), - REG("fscreate", S_IRUGO|S_IWUGO, pid_attr), - REG("keycreate", S_IRUGO|S_IWUGO, pid_attr), - REG("sockcreate", S_IRUGO|S_IWUGO, pid_attr), + REG("current", S_IRUGO|S_IWUGO, proc_pid_attr_operations), + REG("prev", S_IRUGO, proc_pid_attr_operations), + REG("exec", S_IRUGO|S_IWUGO, proc_pid_attr_operations), + REG("fscreate", S_IRUGO|S_IWUGO, proc_pid_attr_operations), + REG("keycreate", S_IRUGO|S_IWUGO, proc_pid_attr_operations), + REG("sockcreate", S_IRUGO|S_IWUGO, proc_pid_attr_operations), }; static int proc_attr_dir_readdir(struct file * filp, @@ -2461,74 +2458,74 @@ static const struct file_operations proc_task_operations; static const struct inode_operations proc_task_inode_operations; static const struct pid_entry tgid_base_stuff[] = { - DIR("task", S_IRUGO|S_IXUGO, task), - DIR("fd", S_IRUSR|S_IXUSR, fd), - DIR("fdinfo", S_IRUSR|S_IXUSR, fdinfo), + DIR("task", S_IRUGO|S_IXUGO, proc_task_inode_operations, proc_task_operations), + DIR("fd", S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations), + DIR("fdinfo", S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations), #ifdef CONFIG_NET - DIR("net", S_IRUGO|S_IXUGO, net), + DIR("net", S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations), #endif - REG("environ", S_IRUSR, environ), - INF("auxv", S_IRUSR, pid_auxv), - ONE("status", S_IRUGO, pid_status), - ONE("personality", S_IRUSR, pid_personality), - INF("limits", S_IRUSR, pid_limits), + REG("environ", S_IRUSR, proc_environ_operations), + INF("auxv", S_IRUSR, proc_pid_auxv), + ONE("status", S_IRUGO, proc_pid_status), + ONE("personality", S_IRUSR, proc_pid_personality), + INF("limits", S_IRUSR, proc_pid_limits), #ifdef CONFIG_SCHED_DEBUG - REG("sched", S_IRUGO|S_IWUSR, pid_sched), + REG("sched", S_IRUGO|S_IWUSR, proc_pid_sched_operations), #endif #ifdef CONFIG_HAVE_ARCH_TRACEHOOK - INF("syscall", S_IRUSR, pid_syscall), + INF("syscall", S_IRUSR, proc_pid_syscall), #endif - INF("cmdline", S_IRUGO, pid_cmdline), - ONE("stat", S_IRUGO, tgid_stat), - ONE("statm", S_IRUGO, pid_statm), - REG("maps", S_IRUGO, maps), + INF("cmdline", S_IRUGO, proc_pid_cmdline), + ONE("stat", S_IRUGO, proc_tgid_stat), + ONE("statm", S_IRUGO, proc_pid_statm), + REG("maps", S_IRUGO, proc_maps_operations), #ifdef CONFIG_NUMA - REG("numa_maps", S_IRUGO, numa_maps), + REG("numa_maps", S_IRUGO, proc_numa_maps_operations), #endif - REG("mem", S_IRUSR|S_IWUSR, mem), - LNK("cwd", cwd), - LNK("root", root), - LNK("exe", exe), - REG("mounts", S_IRUGO, mounts), - REG("mountinfo", S_IRUGO, mountinfo), - REG("mountstats", S_IRUSR, mountstats), + REG("mem", S_IRUSR|S_IWUSR, proc_mem_operations), + LNK("cwd", proc_cwd_link), + LNK("root", proc_root_link), + LNK("exe", proc_exe_link), + REG("mounts", S_IRUGO, proc_mounts_operations), + REG("mountinfo", S_IRUGO, proc_mountinfo_operations), + REG("mountstats", S_IRUSR, proc_mountstats_operations), #ifdef CONFIG_PROC_PAGE_MONITOR - REG("clear_refs", S_IWUSR, clear_refs), - REG("smaps", S_IRUGO, smaps), - REG("pagemap", S_IRUSR, pagemap), + REG("clear_refs", S_IWUSR, proc_clear_refs_operations), + REG("smaps", S_IRUGO, proc_smaps_operations), + REG("pagemap", S_IRUSR, proc_pagemap_operations), #endif #ifdef CONFIG_SECURITY - DIR("attr", S_IRUGO|S_IXUGO, attr_dir), + DIR("attr", S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations), #endif #ifdef CONFIG_KALLSYMS - INF("wchan", S_IRUGO, pid_wchan), + INF("wchan", S_IRUGO, proc_pid_wchan), #endif #ifdef CONFIG_SCHEDSTATS - INF("schedstat", S_IRUGO, pid_schedstat), + INF("schedstat", S_IRUGO, proc_pid_schedstat), #endif #ifdef CONFIG_LATENCYTOP - REG("latency", S_IRUGO, lstats), + REG("latency", S_IRUGO, proc_lstats_operations), #endif #ifdef CONFIG_PROC_PID_CPUSET - REG("cpuset", S_IRUGO, cpuset), + REG("cpuset", S_IRUGO, proc_cpuset_operations), #endif #ifdef CONFIG_CGROUPS - REG("cgroup", S_IRUGO, cgroup), + REG("cgroup", S_IRUGO, proc_cgroup_operations), #endif - INF("oom_score", S_IRUGO, oom_score), - REG("oom_adj", S_IRUGO|S_IWUSR, oom_adjust), + INF("oom_score", S_IRUGO, proc_oom_score), + REG("oom_adj", S_IRUGO|S_IWUSR, proc_oom_adjust_operations), #ifdef CONFIG_AUDITSYSCALL - REG("loginuid", S_IWUSR|S_IRUGO, loginuid), - REG("sessionid", S_IRUGO, sessionid), + REG("loginuid", S_IWUSR|S_IRUGO, proc_loginuid_operations), + REG("sessionid", S_IRUGO, proc_sessionid_operations), #endif #ifdef CONFIG_FAULT_INJECTION - REG("make-it-fail", S_IRUGO|S_IWUSR, fault_inject), + REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations), #endif #if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE) - REG("coredump_filter", S_IRUGO|S_IWUSR, coredump_filter), + REG("coredump_filter", S_IRUGO|S_IWUSR, proc_coredump_filter_operations), #endif #ifdef CONFIG_TASK_IO_ACCOUNTING - INF("io", S_IRUGO, tgid_io_accounting), + INF("io", S_IRUGO, proc_tgid_io_accounting), #endif }; @@ -2801,66 +2798,66 @@ out_no_task: * Tasks */ static const struct pid_entry tid_base_stuff[] = { - DIR("fd", S_IRUSR|S_IXUSR, fd), - DIR("fdinfo", S_IRUSR|S_IXUSR, fdinfo), - REG("environ", S_IRUSR, environ), - INF("auxv", S_IRUSR, pid_auxv), - ONE("status", S_IRUGO, pid_status), - ONE("personality", S_IRUSR, pid_personality), - INF("limits", S_IRUSR, pid_limits), + DIR("fd", S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations), + DIR("fdinfo", S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fd_operations), + REG("environ", S_IRUSR, proc_environ_operations), + INF("auxv", S_IRUSR, proc_pid_auxv), + ONE("status", S_IRUGO, proc_pid_status), + ONE("personality", S_IRUSR, proc_pid_personality), + INF("limits", S_IRUSR, proc_pid_limits), #ifdef CONFIG_SCHED_DEBUG - REG("sched", S_IRUGO|S_IWUSR, pid_sched), + REG("sched", S_IRUGO|S_IWUSR, proc_pid_sched_operations), #endif #ifdef CONFIG_HAVE_ARCH_TRACEHOOK - INF("syscall", S_IRUSR, pid_syscall), + INF("syscall", S_IRUSR, proc_pid_syscall), #endif - INF("cmdline", S_IRUGO, pid_cmdline), - ONE("stat", S_IRUGO, tid_stat), - ONE("statm", S_IRUGO, pid_statm), - REG("maps", S_IRUGO, maps), + INF("cmdline", S_IRUGO, proc_pid_cmdline), + ONE("stat", S_IRUGO, proc_tid_stat), + ONE("statm", S_IRUGO, proc_pid_statm), + REG("maps", S_IRUGO, proc_maps_operations), #ifdef CONFIG_NUMA - REG("numa_maps", S_IRUGO, numa_maps), + REG("numa_maps", S_IRUGO, proc_numa_maps_operations), #endif - REG("mem", S_IRUSR|S_IWUSR, mem), - LNK("cwd", cwd), - LNK("root", root), - LNK("exe", exe), - REG("mounts", S_IRUGO, mounts), - REG("mountinfo", S_IRUGO, mountinfo), + REG("mem", S_IRUSR|S_IWUSR, proc_mem_operations), + LNK("cwd", proc_cwd_link), + LNK("root", proc_root_link), + LNK("exe", proc_exe_link), + REG("mounts", S_IRUGO, proc_mounts_operations), + REG("mountinfo", S_IRUGO, proc_mountinfo_operations), #ifdef CONFIG_PROC_PAGE_MONITOR - REG("clear_refs", S_IWUSR, clear_refs), - REG("smaps", S_IRUGO, smaps), - REG("pagemap", S_IRUSR, pagemap), + REG("clear_refs", S_IWUSR, proc_clear_refs_operations), + REG("smaps", S_IRUGO, proc_smaps_operations), + REG("pagemap", S_IRUSR, proc_pagemap_operations), #endif #ifdef CONFIG_SECURITY - DIR("attr", S_IRUGO|S_IXUGO, attr_dir), + DIR("attr", S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations), #endif #ifdef CONFIG_KALLSYMS - INF("wchan", S_IRUGO, pid_wchan), + INF("wchan", S_IRUGO, proc_pid_wchan), #endif #ifdef CONFIG_SCHEDSTATS - INF("schedstat", S_IRUGO, pid_schedstat), + INF("schedstat", S_IRUGO, proc_pid_schedstat), #endif #ifdef CONFIG_LATENCYTOP - REG("latency", S_IRUGO, lstats), + REG("latency", S_IRUGO, proc_lstats_operations), #endif #ifdef CONFIG_PROC_PID_CPUSET - REG("cpuset", S_IRUGO, cpuset), + REG("cpuset", S_IRUGO, proc_cpuset_operations), #endif #ifdef CONFIG_CGROUPS - REG("cgroup", S_IRUGO, cgroup), + REG("cgroup", S_IRUGO, proc_cgroup_operations), #endif - INF("oom_score", S_IRUGO, oom_score), - REG("oom_adj", S_IRUGO|S_IWUSR, oom_adjust), + INF("oom_score", S_IRUGO, proc_oom_score), + REG("oom_adj", S_IRUGO|S_IWUSR, proc_oom_adjust_operations), #ifdef CONFIG_AUDITSYSCALL - REG("loginuid", S_IWUSR|S_IRUGO, loginuid), - REG("sessionid", S_IRUSR, sessionid), + REG("loginuid", S_IWUSR|S_IRUGO, proc_loginuid_operations), + REG("sessionid", S_IRUSR, proc_sessionid_operations), #endif #ifdef CONFIG_FAULT_INJECTION - REG("make-it-fail", S_IRUGO|S_IWUSR, fault_inject), + REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations), #endif #ifdef CONFIG_TASK_IO_ACCOUNTING - INF("io", S_IRUGO, tid_io_accounting), + INF("io", S_IRUGO, proc_tid_io_accounting), #endif }; -- cgit v1.2.3-70-g09d2 From 2ec220e27f5040aec1e88901c1b6ea3d135787ad Mon Sep 17 00:00:00 2001 From: Ken Chen Date: Mon, 10 Nov 2008 11:26:08 +0300 Subject: proc: add /proc/*/stack /proc/*/stack adds the ability to query a task's stack trace. It is more useful than /proc/*/wchan as it provides full stack trace instead of single depth. Example output: $ cat /proc/self/stack [] save_stack_trace_tsk+0x17/0x35 [] proc_pid_stack+0x4a/0x76 [] proc_single_show+0x4a/0x5e [] seq_read+0xf3/0x29f [] vfs_read+0x6d/0x91 [] sys_read+0x3b/0x60 [] syscall_call+0x7/0xb [] 0xffffffff [add save_stack_trace_tsk() on mips, ACK Ralf --adobriyan] Signed-off-by: Ken Chen Signed-off-by: Ingo Molnar Signed-off-by: Alexey Dobriyan --- Documentation/filesystems/proc.txt | 1 + arch/mips/kernel/stacktrace.c | 24 ++++++++++++++++++------ fs/proc/base.c | 38 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 57 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index 71df353e367..334ef2f983f 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -140,6 +140,7 @@ Table 1-1: Process specific entries in /proc statm Process memory status information status Process status in human readable form wchan If CONFIG_KALLSYMS is set, a pre-decoded wchan + stack Report full stack trace, enable via CONFIG_STACKTRACE smaps Extension based on maps, the rss size for each mapped file .............................................................................. diff --git a/arch/mips/kernel/stacktrace.c b/arch/mips/kernel/stacktrace.c index 0632e2a849c..58f5cd76c8c 100644 --- a/arch/mips/kernel/stacktrace.c +++ b/arch/mips/kernel/stacktrace.c @@ -32,7 +32,8 @@ static void save_raw_context_stack(struct stack_trace *trace, } } -static void save_context_stack(struct stack_trace *trace, struct pt_regs *regs) +static void save_context_stack(struct stack_trace *trace, + struct task_struct *tsk, struct pt_regs *regs) { unsigned long sp = regs->regs[29]; #ifdef CONFIG_KALLSYMS @@ -41,7 +42,7 @@ static void save_context_stack(struct stack_trace *trace, struct pt_regs *regs) if (raw_show_trace || !__kernel_text_address(pc)) { unsigned long stack_page = - (unsigned long)task_stack_page(current); + (unsigned long)task_stack_page(tsk); if (stack_page && sp >= stack_page && sp <= stack_page + THREAD_SIZE - 32) save_raw_context_stack(trace, sp); @@ -54,7 +55,7 @@ static void save_context_stack(struct stack_trace *trace, struct pt_regs *regs) trace->entries[trace->nr_entries++] = pc; if (trace->nr_entries >= trace->max_entries) break; - pc = unwind_stack(current, &sp, pc, &ra); + pc = unwind_stack(tsk, &sp, pc, &ra); } while (pc); #else save_raw_context_stack(trace, sp); @@ -65,13 +66,24 @@ static void save_context_stack(struct stack_trace *trace, struct pt_regs *regs) * Save stack-backtrace addresses into a stack_trace buffer. */ void save_stack_trace(struct stack_trace *trace) +{ + save_stack_trace_tsk(current, trace); +} +EXPORT_SYMBOL_GPL(save_stack_trace); + +void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace) { struct pt_regs dummyregs; struct pt_regs *regs = &dummyregs; WARN_ON(trace->nr_entries || !trace->max_entries); - prepare_frametrace(regs); - save_context_stack(trace, regs); + if (tsk != current) { + regs->regs[29] = tsk->thread.reg29; + regs->regs[31] = 0; + regs->cp0_epc = tsk->thread.reg31; + } else + prepare_frametrace(regs); + save_context_stack(trace, tsk, regs); } -EXPORT_SYMBOL_GPL(save_stack_trace); +EXPORT_SYMBOL_GPL(save_stack_trace_tsk); diff --git a/fs/proc/base.c b/fs/proc/base.c index ce7a6da1b6a..eb7b4654d6a 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -65,6 +65,7 @@ #include #include #include +#include #include #include #include @@ -337,6 +338,37 @@ static int proc_pid_wchan(struct task_struct *task, char *buffer) } #endif /* CONFIG_KALLSYMS */ +#ifdef CONFIG_STACKTRACE + +#define MAX_STACK_TRACE_DEPTH 64 + +static int proc_pid_stack(struct seq_file *m, struct pid_namespace *ns, + struct pid *pid, struct task_struct *task) +{ + struct stack_trace trace; + unsigned long *entries; + int i; + + entries = kmalloc(MAX_STACK_TRACE_DEPTH * sizeof(*entries), GFP_KERNEL); + if (!entries) + return -ENOMEM; + + trace.nr_entries = 0; + trace.max_entries = MAX_STACK_TRACE_DEPTH; + trace.entries = entries; + trace.skip = 0; + save_stack_trace_tsk(task, &trace); + + for (i = 0; i < trace.nr_entries; i++) { + seq_printf(m, "[<%p>] %pS\n", + (void *)entries[i], (void *)entries[i]); + } + kfree(entries); + + return 0; +} +#endif + #ifdef CONFIG_SCHEDSTATS /* * Provides /proc/PID/schedstat @@ -2500,6 +2532,9 @@ static const struct pid_entry tgid_base_stuff[] = { #ifdef CONFIG_KALLSYMS INF("wchan", S_IRUGO, proc_pid_wchan), #endif +#ifdef CONFIG_STACKTRACE + ONE("stack", S_IRUSR, proc_pid_stack), +#endif #ifdef CONFIG_SCHEDSTATS INF("schedstat", S_IRUGO, proc_pid_schedstat), #endif @@ -2835,6 +2870,9 @@ static const struct pid_entry tid_base_stuff[] = { #ifdef CONFIG_KALLSYMS INF("wchan", S_IRUGO, proc_pid_wchan), #endif +#ifdef CONFIG_STACKTRACE + ONE("stack", S_IRUSR, proc_pid_stack), +#endif #ifdef CONFIG_SCHEDSTATS INF("schedstat", S_IRUGO, proc_pid_schedstat), #endif -- cgit v1.2.3-70-g09d2 From dfe6b7d9406c631d697f8bbe1eae5569b808154f Mon Sep 17 00:00:00 2001 From: Hannes Eder Date: Tue, 30 Dec 2008 18:49:13 +0300 Subject: proc: fix sparse warning fs/proc/base.c:312:4: warning: do-while statement is not a compound statement Signed-off-by: Hannes Eder Signed-off-by: Alexey Dobriyan --- fs/proc/base.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index eb7b4654d6a..a9ccc125273 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -306,9 +306,9 @@ static int proc_pid_auxv(struct task_struct *task, char *buffer) struct mm_struct *mm = get_task_mm(task); if (mm) { unsigned int nwords = 0; - do + do { nwords += 2; - while (mm->saved_auxv[nwords - 2] != 0); /* AT_NULL */ + } while (mm->saved_auxv[nwords - 2] != 0); /* AT_NULL */ res = nwords * sizeof(mm->saved_auxv[0]); if (res > PAGE_SIZE) res = PAGE_SIZE; -- cgit v1.2.3-70-g09d2 From 230e40fbda242544389a5428a2efac568178ddfe Mon Sep 17 00:00:00 2001 From: WANG Cong Date: Tue, 30 Dec 2008 19:10:35 +0300 Subject: proc: remove write-only variable in proc_pident_lookup() Signed-off-by: WANG Cong Signed-off-by: Alexey Dobriyan --- fs/proc/base.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index a9ccc125273..0bc9ca03b91 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -2001,13 +2001,11 @@ static struct dentry *proc_pident_lookup(struct inode *dir, const struct pid_entry *ents, unsigned int nents) { - struct inode *inode; struct dentry *error; struct task_struct *task = get_proc_task(dir); const struct pid_entry *p, *last; error = ERR_PTR(-ENOENT); - inode = NULL; if (!task) goto out_no_task; -- cgit v1.2.3-70-g09d2 From 025dfdafe77f20b3890981a394774baab7b9c827 Mon Sep 17 00:00:00 2001 From: Frederik Schwarzer Date: Thu, 16 Oct 2008 19:02:37 +0200 Subject: trivial: fix then -> than typos in comments and documentation - (better, more, bigger ...) then -> (...) than Signed-off-by: Frederik Schwarzer Signed-off-by: Jiri Kosina --- Documentation/hwmon/abituguru-datasheet | 6 +++--- Documentation/networking/rxrpc.txt | 2 +- Documentation/scsi/ChangeLog.lpfc | 2 +- arch/blackfin/kernel/kgdb.c | 2 +- arch/ia64/kernel/kprobes.c | 2 +- arch/m68k/Kconfig | 2 +- arch/mips/pmc-sierra/yosemite/atmel_read_eeprom.c | 2 +- arch/powerpc/kernel/kprobes.c | 2 +- arch/powerpc/oprofile/cell/spu_profiler.c | 2 +- arch/s390/Kconfig | 2 +- arch/s390/kernel/kprobes.c | 2 +- arch/sparc/kernel/kprobes.c | 2 +- arch/x86/kernel/kprobes.c | 2 +- arch/x86/kernel/mfgpt_32.c | 2 +- drivers/hwmon/fschmd.c | 2 +- drivers/infiniband/hw/mlx4/cq.c | 2 +- drivers/message/i2o/i2o_scsi.c | 2 +- drivers/mtd/devices/pmc551.c | 2 +- drivers/mtd/ubi/eba.c | 2 +- drivers/mtd/ubi/io.c | 2 +- drivers/mtd/ubi/scan.c | 2 +- drivers/mtd/ubi/ubi-media.h | 4 ++-- drivers/mtd/ubi/vtbl.c | 2 +- drivers/mtd/ubi/wl.c | 4 ++-- drivers/net/bnx2x_link.c | 2 +- drivers/net/e1000/e1000_hw.c | 4 ++-- drivers/net/slip.h | 2 +- drivers/net/tehuti.c | 4 ++-- drivers/net/tokenring/smctr.c | 2 +- drivers/net/wireless/ipw2x00/ipw2100.c | 2 +- drivers/net/wireless/rt2x00/rt2x00crypto.c | 4 ++-- drivers/net/wireless/strip.c | 2 +- drivers/s390/block/dasd_eer.c | 4 ++-- drivers/s390/char/vmlogrdr.c | 4 ++-- drivers/scsi/lpfc/lpfc_hbadisc.c | 4 ++-- drivers/scsi/lpfc/lpfc_sli.c | 10 +++++----- drivers/serial/crisv10.c | 4 ++-- drivers/video/console/vgacon.c | 2 +- fs/ocfs2/cluster/heartbeat.c | 2 +- fs/proc/task_nommu.c | 2 +- fs/ubifs/Kconfig | 2 +- fs/ubifs/budget.c | 4 ++-- fs/ubifs/gc.c | 2 +- fs/ubifs/journal.c | 2 +- fs/ubifs/shrinker.c | 2 +- fs/xfs/linux-2.6/xfs_super.c | 2 +- include/linux/mtd/mtd.h | 2 +- include/linux/spi/spi.h | 4 ++-- include/mtd/ubi-user.h | 2 +- kernel/pid.c | 2 +- kernel/time/jiffies.c | 2 +- net/sctp/auth.c | 4 ++-- net/sctp/sm_statefuns.c | 6 +++--- net/sctp/socket.c | 2 +- net/sctp/tsnmap.c | 2 +- sound/usb/usx2y/usbusx2y.c | 2 +- 56 files changed, 76 insertions(+), 76 deletions(-) (limited to 'fs') diff --git a/Documentation/hwmon/abituguru-datasheet b/Documentation/hwmon/abituguru-datasheet index aef5a9b3684..4d184f2db0e 100644 --- a/Documentation/hwmon/abituguru-datasheet +++ b/Documentation/hwmon/abituguru-datasheet @@ -74,7 +74,7 @@ a sensor. Notice that some banks have both a read and a write address this is how the uGuru determines if a read from or a write to the bank is taking place, thus when reading you should always use the read address and when writing the -write address. The write address is always one (1) more then the read address. +write address. The write address is always one (1) more than the read address. uGuru ready @@ -224,7 +224,7 @@ Bit 3: Beep if alarm (RW) Bit 4: 1 if alarm cause measured temp is over the warning threshold (R) Bit 5: 1 if alarm cause measured volt is over the max threshold (R) Bit 6: 1 if alarm cause measured volt is under the min threshold (R) -Bit 7: Volt sensor: Shutdown if alarm persist for more then 4 seconds (RW) +Bit 7: Volt sensor: Shutdown if alarm persist for more than 4 seconds (RW) Temp sensor: Shutdown if temp is over the shutdown threshold (RW) * This bit is only honored/used by the uGuru if a temp sensor is connected @@ -293,7 +293,7 @@ Byte 0: Alarm behaviour for the selected sensor. A 1 enables the described behaviour. Bit 0: Give an alarm if measured rpm is under the min threshold (RW) Bit 3: Beep if alarm (RW) -Bit 7: Shutdown if alarm persist for more then 4 seconds (RW) +Bit 7: Shutdown if alarm persist for more than 4 seconds (RW) Byte 1: min threshold (scale as bank 0x26) diff --git a/Documentation/networking/rxrpc.txt b/Documentation/networking/rxrpc.txt index c3669a3fb4a..60d05eb77c6 100644 --- a/Documentation/networking/rxrpc.txt +++ b/Documentation/networking/rxrpc.txt @@ -540,7 +540,7 @@ A client would issue an operation by: MSG_MORE should be set in msghdr::msg_flags on all but the last part of the request. Multiple requests may be made simultaneously. - If a call is intended to go to a destination other then the default + If a call is intended to go to a destination other than the default specified through connect(), then msghdr::msg_name should be set on the first request message of that call. diff --git a/Documentation/scsi/ChangeLog.lpfc b/Documentation/scsi/ChangeLog.lpfc index ae3f962a7cf..ff19a52fe00 100644 --- a/Documentation/scsi/ChangeLog.lpfc +++ b/Documentation/scsi/ChangeLog.lpfc @@ -733,7 +733,7 @@ Changes from 20040920 to 20041018 I/O completion path a little more, especially taking care of fast-pathing the non-error case. Also removes tons of dead members and defines from lpfc_scsi.h - e.g. lpfc_target is down - to nothing more then the lpfc_nodelist pointer. + to nothing more than the lpfc_nodelist pointer. * Added binary sysfs file to issue mbox commands * Replaced #if __BIG_ENDIAN with #if __BIG_ENDIAN_BITFIELD for compatibility with the user space applications. diff --git a/arch/blackfin/kernel/kgdb.c b/arch/blackfin/kernel/kgdb.c index b795a207742..1c5afaeb950 100644 --- a/arch/blackfin/kernel/kgdb.c +++ b/arch/blackfin/kernel/kgdb.c @@ -105,7 +105,7 @@ void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs) * Extracts ebp, esp and eip values understandable by gdb from the values * saved by switch_to. * thread.esp points to ebp. flags and ebp are pushed in switch_to hence esp - * prior to entering switch_to is 8 greater then the value that is saved. + * prior to entering switch_to is 8 greater than the value that is saved. * If switch_to changes, change following code appropriately. */ void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p) diff --git a/arch/ia64/kernel/kprobes.c b/arch/ia64/kernel/kprobes.c index f07688da947..0017b9de2dd 100644 --- a/arch/ia64/kernel/kprobes.c +++ b/arch/ia64/kernel/kprobes.c @@ -434,7 +434,7 @@ int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) /* * It is possible to have multiple instances associated with a given * task either because an multiple functions in the call path - * have a return probe installed on them, and/or more then one return + * have a return probe installed on them, and/or more than one return * return probe was registered for a target function. * * We can handle this because: diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig index c825bde17cb..fb87c08c6b5 100644 --- a/arch/m68k/Kconfig +++ b/arch/m68k/Kconfig @@ -303,7 +303,7 @@ config M68KFPU_EMU_EXTRAPREC correct rounding, the emulator can (often) do the same but this extra calculation can cost quite some time, so you can disable it here. The emulator will then "only" calculate with a 64 bit - mantissa and round slightly incorrect, what is more then enough + mantissa and round slightly incorrect, what is more than enough for normal usage. config M68KFPU_EMU_ONLY diff --git a/arch/mips/pmc-sierra/yosemite/atmel_read_eeprom.c b/arch/mips/pmc-sierra/yosemite/atmel_read_eeprom.c index 97862f45496..caf5e9a0acc 100644 --- a/arch/mips/pmc-sierra/yosemite/atmel_read_eeprom.c +++ b/arch/mips/pmc-sierra/yosemite/atmel_read_eeprom.c @@ -148,7 +148,7 @@ int read_eeprom(char *buffer, int eeprom_size, int size) send_byte(W_HEADER); recv_ack(); - /* EEPROM with size of more then 2K need two byte addressing */ + /* EEPROM with size of more than 2K need two byte addressing */ if (eeprom_size > 2048) { send_byte(0x00); recv_ack(); diff --git a/arch/powerpc/kernel/kprobes.c b/arch/powerpc/kernel/kprobes.c index de79915452c..b29005a5a8f 100644 --- a/arch/powerpc/kernel/kprobes.c +++ b/arch/powerpc/kernel/kprobes.c @@ -316,7 +316,7 @@ static int __kprobes trampoline_probe_handler(struct kprobe *p, /* * It is possible to have multiple instances associated with a given * task either because an multiple functions in the call path - * have a return probe installed on them, and/or more then one return + * have a return probe installed on them, and/or more than one return * return probe was registered for a target function. * * We can handle this because: diff --git a/arch/powerpc/oprofile/cell/spu_profiler.c b/arch/powerpc/oprofile/cell/spu_profiler.c index dd499c3e9da..83faa958b9d 100644 --- a/arch/powerpc/oprofile/cell/spu_profiler.c +++ b/arch/powerpc/oprofile/cell/spu_profiler.c @@ -49,7 +49,7 @@ void set_spu_profiling_frequency(unsigned int freq_khz, unsigned int cycles_rese * of precision. This is close enough for the purpose at hand. * * The value of the timeout should be small enough that the hw - * trace buffer will not get more then about 1/3 full for the + * trace buffer will not get more than about 1/3 full for the * maximum user specified (the LFSR value) hw sampling frequency. * This is to ensure the trace buffer will never fill even if the * kernel thread scheduling varies under a heavy system load. diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 19577aeffd7..a94a3c3ae93 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -299,7 +299,7 @@ config WARN_STACK This option enables the compiler options -mwarn-framesize and -mwarn-dynamicstack. If the compiler supports these options it will generate warnings for function which either use alloca or - create a stack frame bigger then CONFIG_WARN_STACK_SIZE. + create a stack frame bigger than CONFIG_WARN_STACK_SIZE. Say N if you are unsure. diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c index 569079ec4ff..267f6698680 100644 --- a/arch/s390/kernel/kprobes.c +++ b/arch/s390/kernel/kprobes.c @@ -381,7 +381,7 @@ static int __kprobes trampoline_probe_handler(struct kprobe *p, /* * It is possible to have multiple instances associated with a given * task either because an multiple functions in the call path - * have a return probe installed on them, and/or more then one return + * have a return probe installed on them, and/or more than one return * return probe was registered for a target function. * * We can handle this because: diff --git a/arch/sparc/kernel/kprobes.c b/arch/sparc/kernel/kprobes.c index 201a6e547e4..3bc6527c95a 100644 --- a/arch/sparc/kernel/kprobes.c +++ b/arch/sparc/kernel/kprobes.c @@ -517,7 +517,7 @@ int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) /* * It is possible to have multiple instances associated with a given * task either because an multiple functions in the call path - * have a return probe installed on them, and/or more then one return + * have a return probe installed on them, and/or more than one return * return probe was registered for a target function. * * We can handle this because: diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index 6c27679ec6a..a116e6d5726 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -694,7 +694,7 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs) /* * It is possible to have multiple instances associated with a given * task either because multiple functions in the call path have - * return probes installed on them, and/or more then one + * return probes installed on them, and/or more than one * return probe was registered for a target function. * * We can handle this because: diff --git a/arch/x86/kernel/mfgpt_32.c b/arch/x86/kernel/mfgpt_32.c index c12314c9e86..8815f3c7fec 100644 --- a/arch/x86/kernel/mfgpt_32.c +++ b/arch/x86/kernel/mfgpt_32.c @@ -252,7 +252,7 @@ EXPORT_SYMBOL_GPL(geode_mfgpt_alloc_timer); /* * The MFPGT timers on the CS5536 provide us with suitable timers to use * as clock event sources - not as good as a HPET or APIC, but certainly - * better then the PIT. This isn't a general purpose MFGPT driver, but + * better than the PIT. This isn't a general purpose MFGPT driver, but * a simplified one designed specifically to act as a clock event source. * For full details about the MFGPT, please consult the CS5536 data sheet. */ diff --git a/drivers/hwmon/fschmd.c b/drivers/hwmon/fschmd.c index 96717036893..8b2d756595d 100644 --- a/drivers/hwmon/fschmd.c +++ b/drivers/hwmon/fschmd.c @@ -75,7 +75,7 @@ static const u8 FSCHMD_REG_VOLT[3] = { 0x45, 0x42, 0x48 }; /* minimum pwm at which the fan is driven (pwm can by increased depending on the temp. Notice that for the scy some fans share there minimum speed. - Also notice that with the scy the sensor order is different then with the + Also notice that with the scy the sensor order is different than with the other chips, this order was in the 2.4 driver and kept for consistency. */ static const u8 FSCHMD_REG_FAN_MIN[5][6] = { { 0x55, 0x65 }, /* pos */ diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c index a3c5af1d7ec..de5263beab4 100644 --- a/drivers/infiniband/hw/mlx4/cq.c +++ b/drivers/infiniband/hw/mlx4/cq.c @@ -367,7 +367,7 @@ int mlx4_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata) if (err) goto out; } else { - /* Can't be smaller then the number of outstanding CQEs */ + /* Can't be smaller than the number of outstanding CQEs */ outst_cqe = mlx4_ib_get_outstanding_cqes(cq); if (entries < outst_cqe + 1) { err = 0; diff --git a/drivers/message/i2o/i2o_scsi.c b/drivers/message/i2o/i2o_scsi.c index 1bcdbbb9e7d..3d45817e6dc 100644 --- a/drivers/message/i2o/i2o_scsi.c +++ b/drivers/message/i2o/i2o_scsi.c @@ -390,7 +390,7 @@ static int i2o_scsi_reply(struct i2o_controller *c, u32 m, * @i2o_dev: the I2O device which was added * * If a I2O device is added we catch the notification, because I2O classes - * other then SCSI peripheral will not be received through + * other than SCSI peripheral will not be received through * i2o_scsi_probe(). */ static void i2o_scsi_notify_device_add(struct i2o_device *i2o_dev) diff --git a/drivers/mtd/devices/pmc551.c b/drivers/mtd/devices/pmc551.c index d38bca64bb1..d2fd550f7e0 100644 --- a/drivers/mtd/devices/pmc551.c +++ b/drivers/mtd/devices/pmc551.c @@ -34,7 +34,7 @@ * aperture size, not the dram size, and the V370PDC supplies no * other method for memory size discovery. This problem is * mostly only relevant when compiled as a module, as the - * unloading of the module with an aperture size smaller then + * unloading of the module with an aperture size smaller than * the ram will cause the driver to detect the onboard memory * size to be equal to the aperture size when the module is * reloaded. Soooo, to help, the module supports an msize diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c index 048a606cebd..25def348e5b 100644 --- a/drivers/mtd/ubi/eba.c +++ b/drivers/mtd/ubi/eba.c @@ -717,7 +717,7 @@ write_error: * to the real data size, although the @buf buffer has to contain the * alignment. In all other cases, @len has to be aligned. * - * It is prohibited to write more then once to logical eraseblocks of static + * It is prohibited to write more than once to logical eraseblocks of static * volumes. This function returns zero in case of success and a negative error * code in case of failure. */ diff --git a/drivers/mtd/ubi/io.c b/drivers/mtd/ubi/io.c index a74118c0574..fe81039f2a7 100644 --- a/drivers/mtd/ubi/io.c +++ b/drivers/mtd/ubi/io.c @@ -465,7 +465,7 @@ out: * This function synchronously erases physical eraseblock @pnum. If @torture * flag is not zero, the physical eraseblock is checked by means of writing * different patterns to it and reading them back. If the torturing is enabled, - * the physical eraseblock is erased more then once. + * the physical eraseblock is erased more than once. * * This function returns the number of erasures made in case of success, %-EIO * if the erasure failed or the torturing test failed, and other negative error diff --git a/drivers/mtd/ubi/scan.c b/drivers/mtd/ubi/scan.c index 41d47e1cf15..ecde202a5a1 100644 --- a/drivers/mtd/ubi/scan.c +++ b/drivers/mtd/ubi/scan.c @@ -478,7 +478,7 @@ int ubi_scan_add_used(struct ubi_device *ubi, struct ubi_scan_info *si, return 0; } else { /* - * This logical eraseblock is older then the one found + * This logical eraseblock is older than the one found * previously. */ if (cmp_res & 4) diff --git a/drivers/mtd/ubi/ubi-media.h b/drivers/mtd/ubi/ubi-media.h index 2ad94040905..8419fdccc79 100644 --- a/drivers/mtd/ubi/ubi-media.h +++ b/drivers/mtd/ubi/ubi-media.h @@ -135,7 +135,7 @@ enum { * The erase counter header takes 64 bytes and has a plenty of unused space for * future usage. The unused fields are zeroed. The @version field is used to * indicate the version of UBI implementation which is supposed to be able to - * work with this UBI image. If @version is greater then the current UBI + * work with this UBI image. If @version is greater than the current UBI * version, the image is rejected. This may be useful in future if something * is changed radically. This field is duplicated in the volume identifier * header. @@ -187,7 +187,7 @@ struct ubi_ec_hdr { * (sequence number) is used to distinguish between older and newer versions of * logical eraseblocks. * - * There are 2 situations when there may be more then one physical eraseblock + * There are 2 situations when there may be more than one physical eraseblock * corresponding to the same logical eraseblock, i.e., having the same @vol_id * and @lnum values in the volume identifier header. Suppose we have a logical * eraseblock L and it is mapped to the physical eraseblock P. diff --git a/drivers/mtd/ubi/vtbl.c b/drivers/mtd/ubi/vtbl.c index 333c8941552..1afc61e7455 100644 --- a/drivers/mtd/ubi/vtbl.c +++ b/drivers/mtd/ubi/vtbl.c @@ -577,7 +577,7 @@ static int init_volumes(struct ubi_device *ubi, const struct ubi_scan_info *si, if (vtbl[i].flags & UBI_VTBL_AUTORESIZE_FLG) { /* Auto re-size flag may be set only for one volume */ if (ubi->autoresize_vol_id != -1) { - ubi_err("more then one auto-resize volume (%d " + ubi_err("more than one auto-resize volume (%d " "and %d)", ubi->autoresize_vol_id, i); kfree(vol); return -EINVAL; diff --git a/drivers/mtd/ubi/wl.c b/drivers/mtd/ubi/wl.c index 14901cb82c1..891534f8210 100644 --- a/drivers/mtd/ubi/wl.c +++ b/drivers/mtd/ubi/wl.c @@ -128,7 +128,7 @@ * situation when the picked physical eraseblock is constantly erased after the * data is written to it. So, we have a constant which limits the highest erase * counter of the free physical eraseblock to pick. Namely, the WL sub-system - * does not pick eraseblocks with erase counter greater then the lowest erase + * does not pick eraseblocks with erase counter greater than the lowest erase * counter plus %WL_FREE_MAX_DIFF. */ #define WL_FREE_MAX_DIFF (2*UBI_WL_THRESHOLD) @@ -917,7 +917,7 @@ static int ensure_wear_leveling(struct ubi_device *ubi) /* * We schedule wear-leveling only if the difference between the * lowest erase counter of used physical eraseblocks and a high - * erase counter of free physical eraseblocks is greater then + * erase counter of free physical eraseblocks is greater than * %UBI_WL_THRESHOLD. */ e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, u.rb); diff --git a/drivers/net/bnx2x_link.c b/drivers/net/bnx2x_link.c index 67de94f1f30..fefa6ab1306 100644 --- a/drivers/net/bnx2x_link.c +++ b/drivers/net/bnx2x_link.c @@ -3359,7 +3359,7 @@ static u8 bnx2x_format_ver(u32 num, u8 *str, u16 len) u8 shift = 8*4; u8 digit; if (len < 10) { - /* Need more then 10chars for this format */ + /* Need more than 10chars for this format */ *str_ptr = '\0'; return -EINVAL; } diff --git a/drivers/net/e1000/e1000_hw.c b/drivers/net/e1000/e1000_hw.c index d04eef53571..e1a3fc1303e 100644 --- a/drivers/net/e1000/e1000_hw.c +++ b/drivers/net/e1000/e1000_hw.c @@ -6758,7 +6758,7 @@ static s32 e1000_get_cable_length(struct e1000_hw *hw, u16 *min_length, * returns: - E1000_ERR_XXX * E1000_SUCCESS * - * For phy's older then IGP, this function simply reads the polarity bit in the + * For phy's older than IGP, this function simply reads the polarity bit in the * Phy Status register. For IGP phy's, this bit is valid only if link speed is * 10 Mbps. If the link speed is 100 Mbps there is no polarity so this bit will * return 0. If the link speed is 1000 Mbps the polarity status is in the @@ -6834,7 +6834,7 @@ static s32 e1000_check_polarity(struct e1000_hw *hw, * returns: - E1000_ERR_XXX * E1000_SUCCESS * - * For phy's older then IGP, this function reads the Downshift bit in the Phy + * For phy's older than IGP, this function reads the Downshift bit in the Phy * Specific Status register. For IGP phy's, it reads the Downgrade bit in the * Link Health register. In IGP this bit is latched high, so the driver must * read it immediately after link is established. diff --git a/drivers/net/slip.h b/drivers/net/slip.h index 853e0f6ec71..9ea5c11287d 100644 --- a/drivers/net/slip.h +++ b/drivers/net/slip.h @@ -75,7 +75,7 @@ struct slip { unsigned long tx_errors; /* Planned stuff */ unsigned long rx_dropped; /* No memory for skb */ unsigned long tx_dropped; /* When MTU change */ - unsigned long rx_over_errors; /* Frame bigger then SLIP buf. */ + unsigned long rx_over_errors; /* Frame bigger than SLIP buf. */ #ifdef SL_INCLUDE_CSLIP unsigned long tx_compressed; unsigned long rx_compressed; diff --git a/drivers/net/tehuti.c b/drivers/net/tehuti.c index a10a83a11d9..a7a4dc4d631 100644 --- a/drivers/net/tehuti.c +++ b/drivers/net/tehuti.c @@ -1004,7 +1004,7 @@ static inline void bdx_rxdb_free_elem(struct rxdb *db, int n) * skb for rx. It assumes that Rx is desabled in HW * funcs are grouped for better cache usage * - * RxD fifo is smaller then RxF fifo by design. Upon high load, RxD will be + * RxD fifo is smaller than RxF fifo by design. Upon high load, RxD will be * filled and packets will be dropped by nic without getting into host or * cousing interrupt. Anyway, in that condition, host has no chance to proccess * all packets, but dropping in nic is cheaper, since it takes 0 cpu cycles @@ -1826,7 +1826,7 @@ static void bdx_tx_free(struct bdx_priv *priv) * * Pushes desc to TxD fifo and overlaps it if needed. * NOTE: this func does not check for available space. this is responsibility - * of the caller. Neither does it check that data size is smaller then + * of the caller. Neither does it check that data size is smaller than * fifo size. */ static void bdx_tx_push_desc(struct bdx_priv *priv, void *data, int size) diff --git a/drivers/net/tokenring/smctr.c b/drivers/net/tokenring/smctr.c index a011666342f..50eb29ce3c8 100644 --- a/drivers/net/tokenring/smctr.c +++ b/drivers/net/tokenring/smctr.c @@ -3064,7 +3064,7 @@ static int smctr_load_node_addr(struct net_device *dev) * will consequently cause a timeout. * * NOTE 1: If the monitor_state is MS_BEACON_TEST_STATE, all transmit - * queues other then the one used for the lobe_media_test should be + * queues other than the one used for the lobe_media_test should be * disabled.!? * * NOTE 2: If the monitor_state is MS_BEACON_TEST_STATE and the receive_mask diff --git a/drivers/net/wireless/ipw2x00/ipw2100.c b/drivers/net/wireless/ipw2x00/ipw2100.c index 1667065b86a..753de1a9c4b 100644 --- a/drivers/net/wireless/ipw2x00/ipw2100.c +++ b/drivers/net/wireless/ipw2x00/ipw2100.c @@ -1332,7 +1332,7 @@ static int ipw2100_power_cycle_adapter(struct ipw2100_priv *priv) IPW_AUX_HOST_RESET_REG_STOP_MASTER); /* Step 2. Wait for stop Master Assert - * (not more then 50us, otherwise ret error */ + * (not more than 50us, otherwise ret error */ i = 5; do { udelay(IPW_WAIT_RESET_MASTER_ASSERT_COMPLETE_DELAY); diff --git a/drivers/net/wireless/rt2x00/rt2x00crypto.c b/drivers/net/wireless/rt2x00/rt2x00crypto.c index 37ad0d2fb64..aee9cba13eb 100644 --- a/drivers/net/wireless/rt2x00/rt2x00crypto.c +++ b/drivers/net/wireless/rt2x00/rt2x00crypto.c @@ -184,8 +184,8 @@ void rt2x00crypto_rx_insert_iv(struct sk_buff *skb, unsigned int align, * Make room for new data, note that we increase both * headsize and tailsize when required. The tailsize is * only needed when ICV data needs to be inserted and - * the padding is smaller then the ICV data. - * When alignment requirements is greater then the + * the padding is smaller than the ICV data. + * When alignment requirements is greater than the * ICV data we must trim the skb to the correct size * because we need to remove the extra bytes. */ diff --git a/drivers/net/wireless/strip.c b/drivers/net/wireless/strip.c index dd0de3a9ed4..7015f248055 100644 --- a/drivers/net/wireless/strip.c +++ b/drivers/net/wireless/strip.c @@ -236,7 +236,7 @@ struct strip { unsigned long tx_errors; /* Planned stuff */ unsigned long rx_dropped; /* No memory for skb */ unsigned long tx_dropped; /* When MTU change */ - unsigned long rx_over_errors; /* Frame bigger then STRIP buf. */ + unsigned long rx_over_errors; /* Frame bigger than STRIP buf. */ unsigned long pps_timer; /* Timer to determine pps */ unsigned long rx_pps_count; /* Counter to determine pps */ diff --git a/drivers/s390/block/dasd_eer.c b/drivers/s390/block/dasd_eer.c index 892e2878d61..f8e05ce9862 100644 --- a/drivers/s390/block/dasd_eer.c +++ b/drivers/s390/block/dasd_eer.c @@ -535,8 +535,8 @@ static int dasd_eer_open(struct inode *inp, struct file *filp) eerb->buffer_page_count > INT_MAX / PAGE_SIZE) { kfree(eerb); MESSAGE(KERN_WARNING, "can't open device since module " - "parameter eer_pages is smaller then 1 or" - " bigger then %d", (int)(INT_MAX / PAGE_SIZE)); + "parameter eer_pages is smaller than 1 or" + " bigger than %d", (int)(INT_MAX / PAGE_SIZE)); unlock_kernel(); return -EINVAL; } diff --git a/drivers/s390/char/vmlogrdr.c b/drivers/s390/char/vmlogrdr.c index aabbeb909cc..d8a2289fcb6 100644 --- a/drivers/s390/char/vmlogrdr.c +++ b/drivers/s390/char/vmlogrdr.c @@ -427,7 +427,7 @@ static int vmlogrdr_receive_data(struct vmlogrdr_priv_t *priv) buffer = priv->buffer + sizeof(int); } /* - * If the record is bigger then our buffer, we receive only + * If the record is bigger than our buffer, we receive only * a part of it. We can get the rest later. */ if (iucv_data_count > NET_BUFFER_SIZE) @@ -437,7 +437,7 @@ static int vmlogrdr_receive_data(struct vmlogrdr_priv_t *priv) 0, buffer, iucv_data_count, &priv->residual_length); spin_unlock_bh(&priv->priv_lock); - /* An rc of 5 indicates that the record was bigger then + /* An rc of 5 indicates that the record was bigger than * the buffer, which is OK for us. A 9 indicates that the * record was purged befor we could receive it. */ diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c index 8c64494444b..311ed6dea72 100644 --- a/drivers/scsi/lpfc/lpfc_hbadisc.c +++ b/drivers/scsi/lpfc/lpfc_hbadisc.c @@ -1964,10 +1964,10 @@ lpfc_set_disctmo(struct lpfc_vport *vport) uint32_t tmo; if (vport->port_state == LPFC_LOCAL_CFG_LINK) { - /* For FAN, timeout should be greater then edtov */ + /* For FAN, timeout should be greater than edtov */ tmo = (((phba->fc_edtov + 999) / 1000) + 1); } else { - /* Normal discovery timeout should be > then ELS/CT timeout + /* Normal discovery timeout should be > than ELS/CT timeout * FC spec states we need 3 * ratov for CT requests */ tmo = ((phba->fc_ratov * 3) + 3); diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c index 01dfdc8696f..a36a120561e 100644 --- a/drivers/scsi/lpfc/lpfc_sli.c +++ b/drivers/scsi/lpfc/lpfc_sli.c @@ -420,7 +420,7 @@ lpfc_sli_next_iocb_slot (struct lpfc_hba *phba, struct lpfc_sli_ring *pring) if (unlikely(pring->local_getidx >= max_cmd_idx)) { lpfc_printf_log(phba, KERN_ERR, LOG_SLI, "0315 Ring %d issue: portCmdGet %d " - "is bigger then cmd ring %d\n", + "is bigger than cmd ring %d\n", pring->ringno, pring->local_getidx, max_cmd_idx); @@ -1628,12 +1628,12 @@ lpfc_sli_rsp_pointers_error(struct lpfc_hba *phba, struct lpfc_sli_ring *pring) { struct lpfc_pgp *pgp = &phba->port_gp[pring->ringno]; /* - * Ring handler: portRspPut is bigger then + * Ring handler: portRspPut is bigger than * rsp ring */ lpfc_printf_log(phba, KERN_ERR, LOG_SLI, "0312 Ring %d handler: portRspPut %d " - "is bigger then rsp ring %d\n", + "is bigger than rsp ring %d\n", pring->ringno, le32_to_cpu(pgp->rspPutInx), pring->numRiocb); @@ -2083,12 +2083,12 @@ lpfc_sli_handle_slow_ring_event(struct lpfc_hba *phba, portRspPut = le32_to_cpu(pgp->rspPutInx); if (portRspPut >= portRspMax) { /* - * Ring handler: portRspPut is bigger then + * Ring handler: portRspPut is bigger than * rsp ring */ lpfc_printf_log(phba, KERN_ERR, LOG_SLI, "0303 Ring %d handler: portRspPut %d " - "is bigger then rsp ring %d\n", + "is bigger than rsp ring %d\n", pring->ringno, portRspPut, portRspMax); phba->link_state = LPFC_HBA_ERROR; diff --git a/drivers/serial/crisv10.c b/drivers/serial/crisv10.c index 8b2c619a09f..e642c22c80e 100644 --- a/drivers/serial/crisv10.c +++ b/drivers/serial/crisv10.c @@ -1203,7 +1203,7 @@ static void e100_disable_txdma_channel(struct e100_serial *info) unsigned long flags; /* Disable output DMA channel for the serial port in question - * ( set to something other then serialX) + * ( set to something other than serialX) */ local_irq_save(flags); DFLOW(DEBUG_LOG(info->line, "disable_txdma_channel %i\n", info->line)); @@ -1266,7 +1266,7 @@ static void e100_disable_rxdma_channel(struct e100_serial *info) unsigned long flags; /* Disable input DMA channel for the serial port in question - * ( set to something other then serialX) + * ( set to something other than serialX) */ local_irq_save(flags); if (info->line == 0) { diff --git a/drivers/video/console/vgacon.c b/drivers/video/console/vgacon.c index e6210725b9a..d012edda6d1 100644 --- a/drivers/video/console/vgacon.c +++ b/drivers/video/console/vgacon.c @@ -1332,7 +1332,7 @@ static void vgacon_save_screen(struct vc_data *c) c->vc_y = screen_info.orig_y; } - /* We can't copy in more then the size of the video buffer, + /* We can't copy in more than the size of the video buffer, * or we'll be copying in VGA BIOS */ if (!vga_is_gfx) diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index 6ebaa58e2c0..04697ba7f73 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -854,7 +854,7 @@ static int o2hb_thread(void *data) while (!kthread_should_stop() && !reg->hr_unclean_stop) { /* We track the time spent inside - * o2hb_do_disk_heartbeat so that we avoid more then + * o2hb_do_disk_heartbeat so that we avoid more than * hr_timeout_ms between disk writes. On busy systems * this should result in a heartbeat which is less * likely to time itself out. */ diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c index 219bd79ea89..d4a8be32b90 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -9,7 +9,7 @@ /* * Logic: we've got two memory sums for each process, "shared", and - * "non-shared". Shared memory may get counted more then once, for + * "non-shared". Shared memory may get counted more than once, for * each process that owns it. Non-shared memory is counted * accurately. */ diff --git a/fs/ubifs/Kconfig b/fs/ubifs/Kconfig index 91ceeda7e5b..e35b54d5059 100644 --- a/fs/ubifs/Kconfig +++ b/fs/ubifs/Kconfig @@ -40,7 +40,7 @@ config UBIFS_FS_ZLIB depends on UBIFS_FS default y help - Zlib copresses better then LZO but it is slower. Say 'Y' if unsure. + Zlib compresses better than LZO but it is slower. Say 'Y' if unsure. # Debugging-related stuff config UBIFS_FS_DEBUG diff --git a/fs/ubifs/budget.c b/fs/ubifs/budget.c index 0e5e54d8292..175f9c590b7 100644 --- a/fs/ubifs/budget.c +++ b/fs/ubifs/budget.c @@ -142,7 +142,7 @@ static long long get_liability(struct ubifs_info *c) * * This function is called when an operation cannot be budgeted because there * is supposedly no free space. But in most cases there is some free space: - * o budgeting is pessimistic, so it always budgets more then it is actually + * o budgeting is pessimistic, so it always budgets more than it is actually * needed, so shrinking the liability is one way to make free space - the * cached data will take less space then it was budgeted for; * o GC may turn some dark space into free space (budgeting treats dark space @@ -606,7 +606,7 @@ void ubifs_release_budget(struct ubifs_info *c, struct ubifs_budget_req *req) * @c: UBIFS file-system description object * * This function converts budget which was allocated for a new page of data to - * the budget of changing an existing page of data. The latter is smaller then + * the budget of changing an existing page of data. The latter is smaller than * the former, so this function only does simple re-calculation and does not * involve any write-back. */ diff --git a/fs/ubifs/gc.c b/fs/ubifs/gc.c index 0bef6501d58..9832f9abe28 100644 --- a/fs/ubifs/gc.c +++ b/fs/ubifs/gc.c @@ -45,7 +45,7 @@ #define SMALL_NODE_WM UBIFS_MAX_DENT_NODE_SZ /* - * GC may need to move more then one LEB to make progress. The below constants + * GC may need to move more than one LEB to make progress. The below constants * define "soft" and "hard" limits on the number of LEBs the garbage collector * may move. */ diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index 10ae25b7d1d..9b7c54e0cd2 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -191,7 +191,7 @@ again: if (wbuf->lnum != -1 && avail >= len) { /* * Someone else has switched the journal head and we have - * enough space now. This happens when more then one process is + * enough space now. This happens when more than one process is * trying to write to the same journal head at the same time. */ dbg_jnl("return LEB %d back, already have LEB %d:%d", diff --git a/fs/ubifs/shrinker.c b/fs/ubifs/shrinker.c index f248533841a..e7bab52a141 100644 --- a/fs/ubifs/shrinker.c +++ b/fs/ubifs/shrinker.c @@ -151,7 +151,7 @@ static int shrink_tnc(struct ubifs_info *c, int nr, int age, int *contention) * @contention: if any contention, this is set to %1 * * This function walks the list of mounted UBIFS file-systems and frees clean - * znodes which are older then @age, until at least @nr znodes are freed. + * znodes which are older than @age, until at least @nr znodes are freed. * Returns the number of freed znodes. */ static int shrink_tnc_trees(int nr, int age, int *contention) diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 36f6cc703ef..be846d606ae 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -1348,7 +1348,7 @@ xfs_finish_flags( { int ronly = (mp->m_flags & XFS_MOUNT_RDONLY); - /* Fail a mount where the logbuf is smaller then the log stripe */ + /* Fail a mount where the logbuf is smaller than the log stripe */ if (xfs_sb_version_haslogv2(&mp->m_sb)) { if (mp->m_logbsize <= 0 && mp->m_sb.sb_logsunit > XLOG_BIG_RECORD_BSIZE) { diff --git a/include/linux/mtd/mtd.h b/include/linux/mtd/mtd.h index eae26bb6430..64433eb411d 100644 --- a/include/linux/mtd/mtd.h +++ b/include/linux/mtd/mtd.h @@ -83,7 +83,7 @@ typedef enum { * @datbuf: data buffer - if NULL only oob data are read/written * @oobbuf: oob data buffer * - * Note, it is allowed to read more then one OOB area at one go, but not write. + * Note, it is allowed to read more than one OOB area at one go, but not write. * The interface assumes that the OOB write requests program only one page's * OOB area. */ diff --git a/include/linux/spi/spi.h b/include/linux/spi/spi.h index 82229317753..68bb1c501d0 100644 --- a/include/linux/spi/spi.h +++ b/include/linux/spi/spi.h @@ -327,9 +327,9 @@ extern struct spi_master *spi_busnum_to_master(u16 busnum); * @tx_dma: DMA address of tx_buf, if @spi_message.is_dma_mapped * @rx_dma: DMA address of rx_buf, if @spi_message.is_dma_mapped * @len: size of rx and tx buffers (in bytes) - * @speed_hz: Select a speed other then the device default for this + * @speed_hz: Select a speed other than the device default for this * transfer. If 0 the default (from @spi_device) is used. - * @bits_per_word: select a bits_per_word other then the device default + * @bits_per_word: select a bits_per_word other than the device default * for this transfer. If 0 the default (from @spi_device) is used. * @cs_change: affects chipselect after this transfer completes * @delay_usecs: microseconds to delay after this transfer before diff --git a/include/mtd/ubi-user.h b/include/mtd/ubi-user.h index ccdc562e444..2dc2eb2b8e2 100644 --- a/include/mtd/ubi-user.h +++ b/include/mtd/ubi-user.h @@ -253,7 +253,7 @@ struct ubi_mkvol_req { * * Re-sizing is possible for both dynamic and static volumes. But while dynamic * volumes may be re-sized arbitrarily, static volumes cannot be made to be - * smaller then the number of bytes they bear. To arbitrarily shrink a static + * smaller than the number of bytes they bear. To arbitrarily shrink a static * volume, it must be wiped out first (by means of volume update operation with * zero number of bytes). */ diff --git a/kernel/pid.c b/kernel/pid.c index 064e76afa50..af9224cdd6c 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -475,7 +475,7 @@ pid_t task_session_nr_ns(struct task_struct *tsk, struct pid_namespace *ns) EXPORT_SYMBOL(task_session_nr_ns); /* - * Used by proc to find the first pid that is greater then or equal to nr. + * Used by proc to find the first pid that is greater than or equal to nr. * * If there is a pid at nr this function is exactly the same as find_pid_ns. */ diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c index 1ca99557e92..06f197560f3 100644 --- a/kernel/time/jiffies.c +++ b/kernel/time/jiffies.c @@ -45,7 +45,7 @@ * * The value 8 is somewhat carefully chosen, as anything * larger can result in overflows. NSEC_PER_JIFFY grows as - * HZ shrinks, so values greater then 8 overflow 32bits when + * HZ shrinks, so values greater than 8 overflow 32bits when * HZ=100. */ #define JIFFIES_SHIFT 8 diff --git a/net/sctp/auth.c b/net/sctp/auth.c index 52db5f60daa..20c576f530f 100644 --- a/net/sctp/auth.c +++ b/net/sctp/auth.c @@ -141,8 +141,8 @@ void sctp_auth_destroy_keys(struct list_head *keys) /* Compare two byte vectors as numbers. Return values * are: * 0 - vectors are equal - * < 0 - vector 1 is smaller then vector2 - * > 0 - vector 1 is greater then vector2 + * < 0 - vector 1 is smaller than vector2 + * > 0 - vector 1 is greater than vector2 * * Algorithm is: * This is performed by selecting the numerically smaller key vector... diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index 1c4e5d6c29c..3a0cd075914 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -4268,9 +4268,9 @@ nomem: /* * Handle a protocol violation when the chunk length is invalid. - * "Invalid" length is identified as smaller then the minimal length a + * "Invalid" length is identified as smaller than the minimal length a * given chunk can be. For example, a SACK chunk has invalid length - * if it's length is set to be smaller then the size of sctp_sack_chunk_t. + * if its length is set to be smaller than the size of sctp_sack_chunk_t. * * We inform the other end by sending an ABORT with a Protocol Violation * error code. @@ -4300,7 +4300,7 @@ static sctp_disposition_t sctp_sf_violation_chunklen( /* * Handle a protocol violation when the parameter length is invalid. - * "Invalid" length is identified as smaller then the minimal length a + * "Invalid" length is identified as smaller than the minimal length a * given parameter can be. */ static sctp_disposition_t sctp_sf_violation_paramlen( diff --git a/net/sctp/socket.c b/net/sctp/socket.c index b14a8f33e42..ff0a8f88de0 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -2717,7 +2717,7 @@ static int sctp_setsockopt_associnfo(struct sock *sk, char __user *optval, int o paths++; } - /* Only validate asocmaxrxt if we have more then + /* Only validate asocmaxrxt if we have more than * one path/transport. We do this because path * retransmissions are only counted when we have more * then one path. diff --git a/net/sctp/tsnmap.c b/net/sctp/tsnmap.c index 35c73e82553..9bd64565021 100644 --- a/net/sctp/tsnmap.c +++ b/net/sctp/tsnmap.c @@ -227,7 +227,7 @@ void sctp_tsnmap_skip(struct sctp_tsnmap *map, __u32 tsn) */ bitmap_zero(map->tsn_map, map->len); } else { - /* If the gap is smaller then the map size, + /* If the gap is smaller than the map size, * shift the map by 'gap' bits and update further. */ bitmap_shift_right(map->tsn_map, map->tsn_map, gap, map->len); diff --git a/sound/usb/usx2y/usbusx2y.c b/sound/usb/usx2y/usbusx2y.c index ca26c532e77..11639bd72a5 100644 --- a/sound/usb/usx2y/usbusx2y.c +++ b/sound/usb/usx2y/usbusx2y.c @@ -238,7 +238,7 @@ static void i_usX2Y_In04Int(struct urb *urb) send = 0; for (j = 0; j < URBS_AsyncSeq && !err; ++j) if (0 == usX2Y->AS04.urb[j]->status) { - struct us428_p4out *p4out = us428ctls->p4out + send; // FIXME if more then 1 p4out is new, 1 gets lost. + struct us428_p4out *p4out = us428ctls->p4out + send; // FIXME if more than 1 p4out is new, 1 gets lost. usb_fill_bulk_urb(usX2Y->AS04.urb[j], usX2Y->chip.dev, usb_sndbulkpipe(usX2Y->chip.dev, 0x04), &p4out->val.vol, p4out->type == eLT_Light ? sizeof(struct us428_lights) : 5, -- cgit v1.2.3-70-g09d2 From 0211a9c8508b2183e0e539509aad60414f1c3813 Mon Sep 17 00:00:00 2001 From: Frederik Schwarzer Date: Mon, 29 Dec 2008 22:14:56 +0100 Subject: trivial: fix an -> a typos in documentation and comments It is always "an" if there is a vowel _spoken_ (not written). So it is: "an hour" (spoken vowel) but "a uniform" (spoken 'j') Signed-off-by: Frederik Schwarzer Signed-off-by: Jiri Kosina --- Documentation/dell_rbu.txt | 4 ++-- Documentation/laptops/thinkpad-acpi.txt | 2 +- Documentation/networking/tuntap.txt | 2 +- arch/m68k/kernel/traps.c | 2 +- drivers/acpi/executer/exprep.c | 2 +- drivers/acpi/executer/exresolv.c | 2 +- drivers/acpi/executer/exstore.c | 2 +- drivers/acpi/resources/rscreate.c | 2 +- drivers/acpi/utilities/utobject.c | 4 ++-- drivers/char/epca.c | 2 +- drivers/cpufreq/Kconfig | 4 ++-- drivers/input/keyboard/atkbd.c | 2 +- drivers/macintosh/Kconfig | 2 +- drivers/misc/phantom.c | 2 +- fs/ncpfs/ioctl.c | 2 +- include/acpi/acmacros.h | 4 ++-- include/acpi/actypes.h | 2 +- include/linux/ncp_fs.h | 2 +- mm/slub.c | 2 +- sound/oss/aedsp16.c | 2 +- 20 files changed, 24 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/Documentation/dell_rbu.txt b/Documentation/dell_rbu.txt index 2c0d631de0c..c11b931f8f9 100644 --- a/Documentation/dell_rbu.txt +++ b/Documentation/dell_rbu.txt @@ -81,8 +81,8 @@ Until this step is completed the driver cannot be unloaded. Also echoing either mono ,packet or init in to image_type will free up the memory allocated by the driver. -If an user by accident executes steps 1 and 3 above without executing step 2; -it will make the /sys/class/firmware/dell_rbu/ entries to disappear. +If a user by accident executes steps 1 and 3 above without executing step 2; +it will make the /sys/class/firmware/dell_rbu/ entries disappear. The entries can be recreated by doing the following echo init > /sys/devices/platform/dell_rbu/image_type NOTE: echoing init in image_type does not change it original value. diff --git a/Documentation/laptops/thinkpad-acpi.txt b/Documentation/laptops/thinkpad-acpi.txt index 71f0fe1fc1b..898b4987bb8 100644 --- a/Documentation/laptops/thinkpad-acpi.txt +++ b/Documentation/laptops/thinkpad-acpi.txt @@ -1475,7 +1475,7 @@ Sysfs interface changelog: 0x020100: Marker for thinkpad-acpi with hot key NVRAM polling support. If you must, use it to know you should not - start an userspace NVRAM poller (allows to detect when + start a userspace NVRAM poller (allows to detect when NVRAM is compiled out by the user because it is unneeded/undesired in the first place). 0x020101: Marker for thinkpad-acpi with hot key NVRAM polling diff --git a/Documentation/networking/tuntap.txt b/Documentation/networking/tuntap.txt index 839cbb71388..c0aab985bad 100644 --- a/Documentation/networking/tuntap.txt +++ b/Documentation/networking/tuntap.txt @@ -118,7 +118,7 @@ As mentioned above, main purpose of TUN/TAP driver is tunneling. It is used by VTun (http://vtun.sourceforge.net). Another interesting application using TUN/TAP is pipsecd -(http://perso.enst.fr/~beyssac/pipsec/), an userspace IPSec +(http://perso.enst.fr/~beyssac/pipsec/), a userspace IPSec implementation that can use complete kernel routing (unlike FreeS/WAN). 3. How does Virtual network device actually work ? diff --git a/arch/m68k/kernel/traps.c b/arch/m68k/kernel/traps.c index 6d813de2baf..184acc90808 100644 --- a/arch/m68k/kernel/traps.c +++ b/arch/m68k/kernel/traps.c @@ -401,7 +401,7 @@ static inline void do_040writebacks(struct frame *fp) * called from sigreturn(), must ensure userspace code didn't * manipulate exception frame to circumvent protection, then complete * pending writebacks - * we just clear TM2 to turn it into an userspace access + * we just clear TM2 to turn it into a userspace access */ asmlinkage void berr_040cleanup(struct frame *fp) { diff --git a/drivers/acpi/executer/exprep.c b/drivers/acpi/executer/exprep.c index 5d438c32989..a7dc87ecee3 100644 --- a/drivers/acpi/executer/exprep.c +++ b/drivers/acpi/executer/exprep.c @@ -404,7 +404,7 @@ acpi_ex_prep_common_field_object(union acpi_operand_object *obj_desc, * * RETURN: Status * - * DESCRIPTION: Construct an union acpi_operand_object of type def_field and + * DESCRIPTION: Construct a union acpi_operand_object of type def_field and * connect it to the parent Node. * ******************************************************************************/ diff --git a/drivers/acpi/executer/exresolv.c b/drivers/acpi/executer/exresolv.c index 89571b92a52..60e8c47128e 100644 --- a/drivers/acpi/executer/exresolv.c +++ b/drivers/acpi/executer/exresolv.c @@ -146,7 +146,7 @@ acpi_ex_resolve_object_to_value(union acpi_operand_object **stack_ptr, stack_desc = *stack_ptr; - /* This is an union acpi_operand_object */ + /* This is a union acpi_operand_object */ switch (ACPI_GET_OBJECT_TYPE(stack_desc)) { case ACPI_TYPE_LOCAL_REFERENCE: diff --git a/drivers/acpi/executer/exstore.c b/drivers/acpi/executer/exstore.c index 3318df4cbd9..1c118ba78ad 100644 --- a/drivers/acpi/executer/exstore.c +++ b/drivers/acpi/executer/exstore.c @@ -274,7 +274,7 @@ acpi_ex_do_debug_object(union acpi_operand_object *source_desc, * * PARAMETERS: *source_desc - Value to be stored * *dest_desc - Where to store it. Must be an NS node - * or an union acpi_operand_object of type + * or a union acpi_operand_object of type * Reference; * walk_state - Current walk state * diff --git a/drivers/acpi/resources/rscreate.c b/drivers/acpi/resources/rscreate.c index c0bbfa2c419..08b8d73e6ee 100644 --- a/drivers/acpi/resources/rscreate.c +++ b/drivers/acpi/resources/rscreate.c @@ -124,7 +124,7 @@ acpi_rs_create_resource_list(union acpi_operand_object *aml_buffer, * * FUNCTION: acpi_rs_create_pci_routing_table * - * PARAMETERS: package_object - Pointer to an union acpi_operand_object + * PARAMETERS: package_object - Pointer to a union acpi_operand_object * package * output_buffer - Pointer to the user's buffer * diff --git a/drivers/acpi/utilities/utobject.c b/drivers/acpi/utilities/utobject.c index c354e7a42bc..4bef3cfbacc 100644 --- a/drivers/acpi/utilities/utobject.c +++ b/drivers/acpi/utilities/utobject.c @@ -297,7 +297,7 @@ union acpi_operand_object *acpi_ut_create_string_object(acpi_size string_size) * * RETURN: TRUE if object is valid, FALSE otherwise * - * DESCRIPTION: Validate a pointer to be an union acpi_operand_object + * DESCRIPTION: Validate a pointer to be a union acpi_operand_object * ******************************************************************************/ @@ -389,7 +389,7 @@ void acpi_ut_delete_object_desc(union acpi_operand_object *object) { ACPI_FUNCTION_TRACE_PTR(ut_delete_object_desc, object); - /* Object must be an union acpi_operand_object */ + /* Object must be a union acpi_operand_object */ if (ACPI_GET_DESCRIPTOR_TYPE(object) != ACPI_DESC_TYPE_OPERAND) { ACPI_ERROR((AE_INFO, diff --git a/drivers/char/epca.c b/drivers/char/epca.c index 39ad820b235..af7c13ca949 100644 --- a/drivers/char/epca.c +++ b/drivers/char/epca.c @@ -769,7 +769,7 @@ static int pc_open(struct tty_struct *tty, struct file *filp) /* Check status of board configured in system. */ /* - * I check to see if the epca_setup routine detected an user error. It + * I check to see if the epca_setup routine detected a user error. It * might be better to put this in pc_init, but for the moment it goes * here. */ diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig index 5f076aef74f..a8c8d9c19d7 100644 --- a/drivers/cpufreq/Kconfig +++ b/drivers/cpufreq/Kconfig @@ -83,7 +83,7 @@ config CPU_FREQ_DEFAULT_GOV_USERSPACE select CPU_FREQ_GOV_USERSPACE help Use the CPUFreq governor 'userspace' as default. This allows - you to set the CPU frequency manually or when an userspace + you to set the CPU frequency manually or when a userspace program shall be able to set the CPU dynamically without having to enable the userspace governor manually. @@ -138,7 +138,7 @@ config CPU_FREQ_GOV_USERSPACE tristate "'userspace' governor for userspace frequency scaling" help Enable this cpufreq governor when you either want to set the - CPU frequency manually or when an userspace program shall + CPU frequency manually or when a userspace program shall be able to set the CPU dynamically, like on LART . diff --git a/drivers/input/keyboard/atkbd.c b/drivers/input/keyboard/atkbd.c index 379b7ff354e..b9e6bef594a 100644 --- a/drivers/input/keyboard/atkbd.c +++ b/drivers/input/keyboard/atkbd.c @@ -65,7 +65,7 @@ MODULE_PARM_DESC(extra, "Enable extra LEDs and keys on IBM RapidAcces, EzKey and /* * Scancode to keycode tables. These are just the default setting, and - * are loadable via an userland utility. + * are loadable via a userland utility. */ static const unsigned short atkbd_set2_keycode[512] = { diff --git a/drivers/macintosh/Kconfig b/drivers/macintosh/Kconfig index b52659620d5..173cf55c64d 100644 --- a/drivers/macintosh/Kconfig +++ b/drivers/macintosh/Kconfig @@ -138,7 +138,7 @@ config PMAC_BACKLIGHT Say Y here to enable Macintosh specific extensions of the generic backlight code. With this enabled, the brightness keys on older PowerBooks will be enabled so you can change the screen brightness. - Newer models should use an userspace daemon like pbbuttonsd. + Newer models should use a userspace daemon like pbbuttonsd. config PMAC_BACKLIGHT_LEGACY bool "Provide legacy ioctl's on /dev/pmu for the backlight" diff --git a/drivers/misc/phantom.c b/drivers/misc/phantom.c index abdebe34738..fa57b67593a 100644 --- a/drivers/misc/phantom.c +++ b/drivers/misc/phantom.c @@ -6,7 +6,7 @@ * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * - * You need an userspace library to cooperate with this driver. It (and other + * You need a userspace library to cooperate with this driver. It (and other * info) may be obtained here: * http://www.fi.muni.cz/~xslaby/phantom.html * or alternatively, you might use OpenHaptics provided by Sensable. diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index 6d04e050c74..f54360f50a9 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -98,7 +98,7 @@ struct compat_ncp_objectname_ioctl { s32 auth_type; u32 object_name_len; - compat_caddr_t object_name; /* an userspace data, in most cases user name */ + compat_caddr_t object_name; /* a userspace data, in most cases user name */ }; struct compat_ncp_fs_info_v2 { diff --git a/include/acpi/acmacros.h b/include/acpi/acmacros.h index a597207e283..1954c9d1d01 100644 --- a/include/acpi/acmacros.h +++ b/include/acpi/acmacros.h @@ -333,8 +333,8 @@ struct acpi_integer_overlay { #define ACPI_INSERT_BITS(target, mask, source) target = ((target & (~(mask))) | (source & mask)) /* - * An struct acpi_namespace_node can appear in some contexts - * where a pointer to an union acpi_operand_object can also + * A struct acpi_namespace_node can appear in some contexts + * where a pointer to a union acpi_operand_object can also * appear. This macro is used to distinguish them. * * The "Descriptor" field is the first field in both structures. diff --git a/include/acpi/actypes.h b/include/acpi/actypes.h index 7220361790b..8222e8de0d1 100644 --- a/include/acpi/actypes.h +++ b/include/acpi/actypes.h @@ -467,7 +467,7 @@ typedef u32 acpi_object_type; /* * These are special object types that never appear in - * a Namespace node, only in an union acpi_operand_object + * a Namespace node, only in a union acpi_operand_object */ #define ACPI_TYPE_LOCAL_EXTRA 0x1C #define ACPI_TYPE_LOCAL_DATA 0x1D diff --git a/include/linux/ncp_fs.h b/include/linux/ncp_fs.h index 9f2d76347f1..f69e66d151c 100644 --- a/include/linux/ncp_fs.h +++ b/include/linux/ncp_fs.h @@ -87,7 +87,7 @@ struct ncp_objectname_ioctl #define NCP_AUTH_NDS 0x32 int auth_type; size_t object_name_len; - void __user * object_name; /* an userspace data, in most cases user name */ + void __user * object_name; /* a userspace data, in most cases user name */ }; struct ncp_privatedata_ioctl diff --git a/mm/slub.c b/mm/slub.c index f0e2892fe40..6392ae5cc6b 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2254,7 +2254,7 @@ static int calculate_sizes(struct kmem_cache *s, int forced_order) * Add some empty padding so that we can catch * overwrites from earlier objects rather than let * tracking information or the free pointer be - * corrupted if an user writes before the start + * corrupted if a user writes before the start * of the object. */ size += sizeof(void *); diff --git a/sound/oss/aedsp16.c b/sound/oss/aedsp16.c index a0274f3dac0..3ee9900ffd7 100644 --- a/sound/oss/aedsp16.c +++ b/sound/oss/aedsp16.c @@ -157,7 +157,7 @@ Started Fri Mar 17 16:13:18 MET 1995 - v0.1 (ALPHA, was an user-level program called AudioExcelDSP16.c) + v0.1 (ALPHA, was a user-level program called AudioExcelDSP16.c) - Initial code. v0.2 (ALPHA) - Cleanups. -- cgit v1.2.3-70-g09d2 From b3d47676d474ecd914c72049c87e71e5f0ffe040 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 20 Oct 2008 13:01:59 -0400 Subject: nfsd: update fh_verify description Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsfh.c | 30 ++++++++++++++++++++++++------ 1 file changed, 24 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index f0da7d9c3a9..019a8a20184 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -258,14 +258,32 @@ out: return error; } -/* - * Perform sanity checks on the dentry in a client's file handle. +/** + * fh_verify - filehandle lookup and access checking + * @rqstp: pointer to current rpc request + * @fhp: filehandle to be verified + * @type: expected type of object pointed to by filehandle + * @access: type of access needed to object + * + * Look up a dentry from the on-the-wire filehandle, check the client's + * access to the export, and set the current task's credentials. + * + * Regardless of success or failure of fh_verify(), fh_put() should be + * called on @fhp when the caller is finished with the filehandle. + * + * fh_verify() may be called multiple times on a given filehandle, for + * example, when processing an NFSv4 compound. The first call will look + * up a dentry using the on-the-wire filehandle. Subsequent calls will + * skip the lookup and just perform the other checks and possibly change + * the current task's credentials. * - * Note that the file handle dentry may need to be freed even after - * an error return. + * @type specifies the type of object expected using one of the S_IF* + * constants defined in include/linux/stat.h. The caller may use zero + * to indicate that it doesn't care, or a negative integer to indicate + * that it expects something not of the given type. * - * This is only called at the start of an nfsproc call, so fhp points to - * a svc_fh which is all 0 except for the over-the-wire file handle. + * @access is formed from the NFSD_MAY_* constants defined in + * include/linux/nfsd/nfsd.h. */ __be32 fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, int access) -- cgit v1.2.3-70-g09d2 From 9346eff0dea1e5855fba25c9fe639d92a4db3135 Mon Sep 17 00:00:00 2001 From: Krishna Kumar Date: Mon, 20 Oct 2008 11:44:28 +0530 Subject: nfsd: Minor cleanup of find_stateid Minor cleanup/rewrite of find_stateid. Compile tested. Signed-off-by: Krishna Kumar Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 13e0e074dbb..06b89df9221 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2435,13 +2435,13 @@ static struct list_head lockstateid_hashtbl[STATEID_HASH_SIZE]; static struct nfs4_stateid * find_stateid(stateid_t *stid, int flags) { - struct nfs4_stateid *local = NULL; + struct nfs4_stateid *local; u32 st_id = stid->si_stateownerid; u32 f_id = stid->si_fileid; unsigned int hashval; dprintk("NFSD: find_stateid flags 0x%x\n",flags); - if ((flags & LOCK_STATE) || (flags & RD_STATE) || (flags & WR_STATE)) { + if (flags & (LOCK_STATE | RD_STATE | WR_STATE)) { hashval = stateid_hashval(st_id, f_id); list_for_each_entry(local, &lockstateid_hashtbl[hashval], st_hash) { if ((local->st_stateid.si_stateownerid == st_id) && @@ -2449,7 +2449,8 @@ find_stateid(stateid_t *stid, int flags) return local; } } - if ((flags & OPEN_STATE) || (flags & RD_STATE) || (flags & WR_STATE)) { + + if (flags & (OPEN_STATE | RD_STATE | WR_STATE)) { hashval = stateid_hashval(st_id, f_id); list_for_each_entry(local, &stateid_hashtbl[hashval], st_hash) { if ((local->st_stateid.si_stateownerid == st_id) && -- cgit v1.2.3-70-g09d2 From 2bd9e7b62e6e1da3f881c40c73d93e9a212ce6de Mon Sep 17 00:00:00 2001 From: Krishna Kumar Date: Mon, 20 Oct 2008 11:47:09 +0530 Subject: nfsd: Fix leaked memory in nfs4_make_rec_clidname cksum.data is not freed up in one error case. Compile tested. Signed-off-by: Krishna Kumar Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4recover.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index 0f9d6efaa62..74f7b67567f 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -116,9 +116,9 @@ nfs4_make_rec_clidname(char *dname, struct xdr_netobj *clname) md5_to_hex(dname, cksum.data); - kfree(cksum.data); status = nfs_ok; out: + kfree(cksum.data); crypto_free_hash(desc.tfm); out_no_tfm: return status; -- cgit v1.2.3-70-g09d2 From c72a476b4b7ecadb80185de31236edb303c1a5d0 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 20 Oct 2008 11:51:58 -0400 Subject: lockd: set svc_serv->sv_maxconn to a more reasonable value (try #3) The default method for calculating the number of connections allowed per RPC service arbitrarily limits single-threaded services to 80 connections. This is too low for services like lockd and artificially limits the number of TCP clients that it can support. Have lockd set a default sv_maxconn value to 1024 (which is the typical default value for RLIMIT_NOFILE. Also add a module parameter to allow an admin to set this to an arbitrary value. Signed-off-by: Jeff Layton Acked-by: Neil Brown Signed-off-by: J. Bruce Fields --- fs/lockd/svc.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 252d80163d0..bc3c3cb62db 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -62,6 +62,9 @@ static unsigned long nlm_timeout = LOCKD_DFLT_TIMEO; static int nlm_udpport, nlm_tcpport; int nsm_use_hostnames = 0; +/* RLIM_NOFILE defaults to 1024. That seems like a reasonable default here. */ +static unsigned int nlm_max_connections = 1024; + /* * Constants needed for the sysctl interface. */ @@ -143,6 +146,9 @@ lockd(void *vrqstp) long timeout = MAX_SCHEDULE_TIMEOUT; RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]); + /* update sv_maxconn if it has changed */ + rqstp->rq_server->sv_maxconn = nlm_max_connections; + if (signalled()) { flush_signals(current); if (nlmsvc_ops) { @@ -276,6 +282,7 @@ int lockd_up(void) } svc_sock_update_bufs(serv); + serv->sv_maxconn = nlm_max_connections; nlmsvc_task = kthread_run(lockd, nlmsvc_rqst, serv->sv_name); if (IS_ERR(nlmsvc_task)) { @@ -485,6 +492,7 @@ module_param_call(nlm_udpport, param_set_port, param_get_int, module_param_call(nlm_tcpport, param_set_port, param_get_int, &nlm_tcpport, 0644); module_param(nsm_use_hostnames, bool, 0644); +module_param(nlm_max_connections, uint, 0644); /* * Initialising and terminating the module. -- cgit v1.2.3-70-g09d2 From 1df40b609ad5a622904eb652109c287fe9c93ec5 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:19:53 -0500 Subject: NLM: Remove address eye-catcher buffers from nlm_host The h_name field in struct nlm_host is a just copy of h_nsmhandle->sm_name. Likewise, the contents of the h_addrbuf field should be identical to the sm_addrbuf field. The h_srcaddrbuf field is used only in one place for debugging. We can live without this until we get %pI formatting for printk(). Currently these buffers are 48 bytes, but we need to support scope IDs in IPv6 presentation addresses, which means making the buffers even larger. Instead, let's find ways to eliminate them to save space. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 10 +++------- include/linux/lockd/lockd.h | 4 +--- 2 files changed, 4 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index abdebf76b82..33bf67af7ab 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -206,6 +206,7 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) goto out; } host->h_name = nsm->sm_name; + host->h_addrbuf = nsm->sm_addrbuf; memcpy(nlm_addr(host), ni->sap, ni->salen); host->h_addrlen = ni->salen; nlm_clear_port(nlm_addr(host)); @@ -232,11 +233,6 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) nrhosts++; - nlm_display_address((struct sockaddr *)&host->h_addr, - host->h_addrbuf, sizeof(host->h_addrbuf)); - nlm_display_address((struct sockaddr *)&host->h_srcaddr, - host->h_srcaddrbuf, sizeof(host->h_srcaddrbuf)); - dprintk("lockd: nlm_lookup_host created host %s\n", host->h_name); @@ -378,8 +374,8 @@ nlm_bind_host(struct nlm_host *host) { struct rpc_clnt *clnt; - dprintk("lockd: nlm_bind_host %s (%s), my addr=%s\n", - host->h_name, host->h_addrbuf, host->h_srcaddrbuf); + dprintk("lockd: nlm_bind_host %s (%s)\n", + host->h_name, host->h_addrbuf); /* Lock host handle */ mutex_lock(&host->h_mutex); diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 3dbdd353156..dae22cb4c38 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -65,9 +65,7 @@ struct nlm_host { struct list_head h_granted; /* Locks in GRANTED state */ struct list_head h_reclaim; /* Locks in RECLAIM state */ struct nsm_handle *h_nsmhandle; /* NSM status handle */ - - char h_addrbuf[48], /* address eyecatchers */ - h_srcaddrbuf[48]; + char *h_addrbuf; /* address eyecatcher */ }; struct nsm_handle { -- cgit v1.2.3-70-g09d2 From 6999fb4016b2604c2f8a65586bba4a62a4b24ce7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:01 -0500 Subject: NLM: Remove AF_UNSPEC arm in nlm_display_address() AF_UNSPEC support is no longer needed in nlm_display_address() now that a presentation address is no longer generated for the h_srcaddr field. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 33bf67af7ab..beb5da81016 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -112,9 +112,6 @@ static void nlm_display_address(const struct sockaddr *sap, const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; switch (sap->sa_family) { - case AF_UNSPEC: - snprintf(buf, len, "unspecified"); - break; case AF_INET: snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); break; -- cgit v1.2.3-70-g09d2 From bc995801a09d1fead0bec1356bfd836911c8eed7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:08 -0500 Subject: NLM: Support IPv6 scope IDs in nlm_display_address() Scope ID support is needed since the kernel's NSM implementation is about to use these displayed addresses as a mon_name in some cases. When nsm_use_hostnames is zero, without scope ID support NSM will fail to handle peers that contact us via a link-local address. Link-local addresses do not work without an interface ID, which is stored in the sockaddr's sin6_scope_id field. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 21 +++++++++++++++------ include/linux/lockd/lockd.h | 10 +++++++++- 2 files changed, 24 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index beb5da81016..012e49aaecd 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -105,22 +105,31 @@ static void nlm_clear_port(struct sockaddr *sap) } } +static void nlm_display_ipv6_address(const struct sockaddr *sap, char *buf, + const size_t len) +{ + const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; + + if (ipv6_addr_v4mapped(&sin6->sin6_addr)) + snprintf(buf, len, "%pI4", &sin6->sin6_addr.s6_addr32[3]); + else if (sin6->sin6_scope_id != 0) + snprintf(buf, len, "%pI6%%%u", &sin6->sin6_addr, + sin6->sin6_scope_id); + else + snprintf(buf, len, "%pI6", &sin6->sin6_addr); +} + static void nlm_display_address(const struct sockaddr *sap, char *buf, const size_t len) { const struct sockaddr_in *sin = (struct sockaddr_in *)sap; - const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; switch (sap->sa_family) { case AF_INET: snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); break; case AF_INET6: - if (ipv6_addr_v4mapped(&sin6->sin6_addr)) - snprintf(buf, len, "%pI4", - &sin6->sin6_addr.s6_addr32[3]); - else - snprintf(buf, len, "%pI6", &sin6->sin6_addr); + nlm_display_ipv6_address(sap, buf, len); break; default: snprintf(buf, len, "unsupported address family"); diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index dae22cb4c38..80a0a2cff2b 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -68,6 +68,14 @@ struct nlm_host { char *h_addrbuf; /* address eyecatcher */ }; +/* + * The largest string sm_addrbuf should hold is a full-size IPv6 address + * (no "::" anywhere) with a scope ID. The buffer size is computed to + * hold eight groups of colon-separated four-hex-digit numbers, a + * percent sign, a scope id (at most 32 bits, in decimal), and NUL. + */ +#define NSM_ADDRBUF ((8 * 4 + 7) + (1 + 10) + 1) + struct nsm_handle { struct list_head sm_link; atomic_t sm_count; @@ -76,7 +84,7 @@ struct nsm_handle { size_t sm_addrlen; unsigned int sm_monitored : 1, sm_sticky : 1; /* don't unmonitor */ - char sm_addrbuf[48]; /* address eyecatcher */ + char sm_addrbuf[NSM_ADDRBUF]; }; /* -- cgit v1.2.3-70-g09d2 From afb03699dc0a920aed3322ad0e6895533941fb1e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:16 -0500 Subject: NLM: Add helper to handle IPv4 addresses Clean up: introduce a helper function to generate IPv4 addresses using the same style as the IPv6 helper function we just added. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 012e49aaecd..780918acd6f 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -105,6 +105,13 @@ static void nlm_clear_port(struct sockaddr *sap) } } +static void nlm_display_ipv4_address(const struct sockaddr *sap, char *buf, + const size_t len) +{ + const struct sockaddr_in *sin = (struct sockaddr_in *)sap; + snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); +} + static void nlm_display_ipv6_address(const struct sockaddr *sap, char *buf, const size_t len) { @@ -122,11 +129,9 @@ static void nlm_display_ipv6_address(const struct sockaddr *sap, char *buf, static void nlm_display_address(const struct sockaddr *sap, char *buf, const size_t len) { - const struct sockaddr_in *sin = (struct sockaddr_in *)sap; - switch (sap->sa_family) { case AF_INET: - snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); + nlm_display_ipv4_address(sap, buf, len); break; case AF_INET6: nlm_display_ipv6_address(sap, buf, len); -- cgit v1.2.3-70-g09d2 From a4846750f090702e2fb848ac4fe5827bcef34060 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:23 -0500 Subject: NSM: Use C99 structure initializer to initialize nsm_args Clean up: Use a C99 structure initializer instead of open-coding the initialization of nsm_args. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index ffd3461f75e..6f6ff410341 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -37,7 +37,13 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) { struct rpc_clnt *clnt; int status; - struct nsm_args args; + struct nsm_args args = { + .addr = nsm_addr_in(nsm)->sin_addr.s_addr, + .prog = NLM_PROGRAM, + .vers = 3, + .proc = NLMPROC_NSM_NOTIFY, + .mon_name = nsm->sm_name, + }; struct rpc_message msg = { .rpc_argp = &args, .rpc_resp = res, @@ -49,12 +55,6 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) goto out; } - memset(&args, 0, sizeof(args)); - args.mon_name = nsm->sm_name; - args.addr = nsm_addr_in(nsm)->sin_addr.s_addr; - args.prog = NLM_PROGRAM; - args.vers = 3; - args.proc = NLMPROC_NSM_NOTIFY; memset(res, 0, sizeof(*res)); msg.rpc_proc = &clnt->cl_procinfo[proc]; -- cgit v1.2.3-70-g09d2 From 5acf43155d1bcc412d892c73f64044f9a826cde6 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:31 -0500 Subject: NSM: convert printk(KERN_DEBUG) to a dprintk() Clean up: make the printk(KERN_DEBUG) in nsm_mon_unmon() a dprintk, and add another dprintk to note if creating an RPC client for the upcall failed. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 6f6ff410341..497dfea02e8 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -52,6 +52,8 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) clnt = nsm_create(); if (IS_ERR(clnt)) { status = PTR_ERR(clnt); + dprintk("lockd: failed to create NSM upcall transport, " + "status=%d\n", status); goto out; } @@ -60,8 +62,8 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) msg.rpc_proc = &clnt->cl_procinfo[proc]; status = rpc_call_sync(clnt, &msg, 0); if (status < 0) - printk(KERN_DEBUG "nsm_mon_unmon: rpc failed, status=%d\n", - status); + dprintk("lockd: NSM upcall RPC failed, status=%d\n", + status); else status = 0; rpc_shutdown_client(clnt); -- cgit v1.2.3-70-g09d2 From 29ed1407ed81086b778ebf12145b048ac3f7e10e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:46 -0500 Subject: NSM: Support IPv6 version of mon_name The "mon_name" argument of the NSMPROC_MON and NSMPROC_UNMON upcalls is a string that contains the hostname or IP address of the remote peer to be notified when this host has rebooted. The sm-notify command uses this identifier to contact the peer when we reboot, so it must be either a well-qualified DNS hostname or a presentation format IP address string. When the "nsm_use_hostnames" sysctl is set to zero, the kernel's NSM provides a presentation format IP address in the "mon_name" argument. Otherwise, the "caller_name" argument from NLM requests is used, which is usually just the DNS hostname of the peer. To support IPv6 addresses for the mon_name argument, we use the nsm_handle's address eye-catcher, which already contains an appropriate presentation format address string. Using the eye-catcher string obviates the need to use a large buffer on the stack to form the presentation address string for the upcall. This patch also addresses a subtle bug. An NSMPROC_MON request and the subsequent NSMPROC_UNMON request for the same peer are required to use the same value for the "mon_name" argument. Otherwise, rpc.statd's NSMPROC_UNMON processing cannot locate the database entry for that peer and remove it. If the setting of nsm_use_hostnames is changed between the time the kernel sends an NSMPROC_MON request and the time it sends the NSMPROC_UNMON request for the same peer, the "mon_name" argument for these two requests may not be the same. This is because the value of "mon_name" is currently chosen at the moment the call is made based on the setting of nsm_use_hostnames To ensure both requests pass identical contents in the "mon_name" argument, we now select which string to use for the argument in the nsm_monitor() function. A pointer to this string is saved in the nsm_handle so it can be used for a subsequent NSMPROC_UNMON upcall. NB: There are other potential problems, such as how nlm_host_rebooted() might behave if nsm_use_hostnames were changed while hosts are still being monitored. This patch does not attempt to address those problems. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 27 ++++++++------------------- include/linux/lockd/lockd.h | 1 + 2 files changed, 9 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 497dfea02e8..a606fbbf804 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -18,8 +18,6 @@ #define NLMDBG_FACILITY NLMDBG_MONITOR -#define XDR_ADDRBUF_LEN (20) - static struct rpc_clnt * nsm_create(void); static struct rpc_program nsm_program; @@ -42,7 +40,7 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) .prog = NLM_PROGRAM, .vers = 3, .proc = NLMPROC_NSM_NOTIFY, - .mon_name = nsm->sm_name, + .mon_name = nsm->sm_mon_name, }; struct rpc_message msg = { .rpc_argp = &args, @@ -87,6 +85,12 @@ nsm_monitor(struct nlm_host *host) if (nsm->sm_monitored) return 0; + /* + * Choose whether to record the caller_name or IP address of + * this peer in the local rpc.statd's database. + */ + nsm->sm_mon_name = nsm_use_hostnames ? nsm->sm_name : nsm->sm_addrbuf; + status = nsm_mon_unmon(nsm, SM_MON, &res); if (status < 0 || res.status != 0) @@ -167,25 +171,10 @@ static __be32 *xdr_encode_nsm_string(__be32 *p, char *string) /* * "mon_name" specifies the host to be monitored. - * - * Linux uses a text version of the IP address of the remote - * host as the host identifier (the "mon_name" argument). - * - * Linux statd always looks up the canonical hostname first for - * whatever remote hostname it receives, so this works alright. */ static __be32 *xdr_encode_mon_name(__be32 *p, struct nsm_args *argp) { - char buffer[XDR_ADDRBUF_LEN + 1]; - char *name = argp->mon_name; - - if (!nsm_use_hostnames) { - snprintf(buffer, XDR_ADDRBUF_LEN, - "%pI4", &argp->addr); - name = buffer; - } - - return xdr_encode_nsm_string(p, name); + return xdr_encode_nsm_string(p, argp->mon_name); } /* diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 54dbb458e73..d3c7247d23e 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -79,6 +79,7 @@ struct nlm_host { struct nsm_handle { struct list_head sm_link; atomic_t sm_count; + char *sm_mon_name; char *sm_name; struct sockaddr_storage sm_addr; size_t sm_addrlen; -- cgit v1.2.3-70-g09d2 From 9fee49024ed19d849413df4ab6ec1a1a60aaae94 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:20:53 -0500 Subject: NSM: Use sm_name instead of h_name in nsm_monitor() and nsm_unmonitor() Clean up: Use the sm_name field for reporting the hostname in nsm_monitor() and nsm_unmonitor(), just as the other functions in fs/lockd/mon.c do. The h_name field is just a copy of the sm_name pointer. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index a606fbbf804..697bdcdd20c 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -79,7 +79,7 @@ nsm_monitor(struct nlm_host *host) struct nsm_res res; int status; - dprintk("lockd: nsm_monitor(%s)\n", host->h_name); + dprintk("lockd: nsm_monitor(%s)\n", nsm->sm_name); BUG_ON(nsm == NULL); if (nsm->sm_monitored) @@ -94,7 +94,7 @@ nsm_monitor(struct nlm_host *host) status = nsm_mon_unmon(nsm, SM_MON, &res); if (status < 0 || res.status != 0) - printk(KERN_NOTICE "lockd: cannot monitor %s\n", host->h_name); + printk(KERN_NOTICE "lockd: cannot monitor %s\n", nsm->sm_name); else nsm->sm_monitored = 1; return status; @@ -116,12 +116,12 @@ nsm_unmonitor(struct nlm_host *host) if (atomic_read(&nsm->sm_count) == 1 && nsm->sm_monitored && !nsm->sm_sticky) { - dprintk("lockd: nsm_unmonitor(%s)\n", host->h_name); + dprintk("lockd: nsm_unmonitor(%s)\n", nsm->sm_name); status = nsm_mon_unmon(nsm, SM_UNMON, &res); if (status < 0) printk(KERN_NOTICE "lockd: cannot unmonitor %s\n", - host->h_name); + nsm->sm_name); else nsm->sm_monitored = 0; } -- cgit v1.2.3-70-g09d2 From 501c1ed3fb5c2648ba1709282c71617910917f66 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:01 -0500 Subject: NLM: Remove redundant printk() in nlmclnt_lock() The nsm_monitor() function already generates a printk(KERN_NOTICE) if the SM_MON upcall fails, so the similar printk() in the nlmclnt_lock() function is redundant. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/clntproc.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 31668b690e0..5ce42e0ed4a 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -518,11 +518,9 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl) unsigned char fl_type; int status = -ENOLCK; - if (nsm_monitor(host) < 0) { - printk(KERN_NOTICE "lockd: failed to monitor %s\n", - host->h_name); + if (nsm_monitor(host) < 0) goto out; - } + fl->fl_flags |= FL_ACCESS; status = do_vfs_lock(fl); fl->fl_flags = fl_flags; -- cgit v1.2.3-70-g09d2 From 5bc74bef7c9b652f0f2aa9c5a8d5ac86881aba79 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:08 -0500 Subject: NSM: Remove BUG_ON() in nsm_monitor() Clean up: Remove the BUG_ON() invocation in nsm_monitor(). It's not likely that nsm_monitor() is ever called with a NULL host pointer, and the code will die anyway if host is NULL. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 697bdcdd20c..bb5fc1bb37f 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -80,7 +80,6 @@ nsm_monitor(struct nlm_host *host) int status; dprintk("lockd: nsm_monitor(%s)\n", nsm->sm_name); - BUG_ON(nsm == NULL); if (nsm->sm_monitored) return 0; -- cgit v1.2.3-70-g09d2 From 5d254b119823658cc318f88589c6c426b3d0a153 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:15 -0500 Subject: NSM: Make sure to return an error if the SM_MON call result is not zero The nsm_monitor() function reports an error and does not set sm_monitored if the SM_MON upcall reply has a non-zero result code, but nsm_monitor() does not return an error to its caller in this case. Since sm_monitored is not set, the upcall is retried when the next NLM request invokes nsm_monitor(). However, that may not come for a while. In the meantime, at least one NLM request will potentially proceed without the peer being monitored properly. Have nsm_monitor() return an error if the result code is non-zero. This will cause all NLM requests to fail immediately if the upcall completed successfully but rpc.statd returned an error. This may be inconvenient in some cases (for example if rpc.statd cannot complete a proper DNS reverse lookup of the hostname), but will make the reboot monitoring service more robust by forcing such issues to be corrected by an admin. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index bb5fc1bb37f..07e16b81498 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -91,8 +91,9 @@ nsm_monitor(struct nlm_host *host) nsm->sm_mon_name = nsm_use_hostnames ? nsm->sm_name : nsm->sm_addrbuf; status = nsm_mon_unmon(nsm, SM_MON, &res); - - if (status < 0 || res.status != 0) + if (res.status != 0) + status = -EIO; + if (status < 0) printk(KERN_NOTICE "lockd: cannot monitor %s\n", nsm->sm_name); else nsm->sm_monitored = 1; -- cgit v1.2.3-70-g09d2 From 1e49323c4ab044d05bbc68cf13cadcbd4372468c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:24 -0500 Subject: NLM: Move the public declaration of nsm_monitor() to lockd.h Clean up. Make the nlm_host argument "const," and move the public declaration to lockd.h with other NSM public function (nsm_release, eg) and global variable declarations. Add a documenting comment. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 15 +++++++++++---- include/linux/lockd/lockd.h | 4 ++++ include/linux/lockd/sm_inter.h | 1 - 3 files changed, 15 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 07e16b81498..aaaa08e7ae7 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -69,11 +69,18 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) return status; } -/* - * Set up monitoring of a remote host +/** + * nsm_monitor - Notify a peer in case we reboot + * @host: pointer to nlm_host of peer to notify + * + * If this peer is not already monitored, this function sends an + * upcall to the local rpc.statd to record the name/address of + * the peer to notify in case we reboot. + * + * Returns zero if the peer is monitored by the local rpc.statd; + * otherwise a negative errno value is returned. */ -int -nsm_monitor(struct nlm_host *host) +int nsm_monitor(const struct nlm_host *host) { struct nsm_handle *nsm = host->h_nsmhandle; struct nsm_res res; diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index d3c7247d23e..f15a4f5ccbf 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -242,6 +242,10 @@ extern void nlm_host_rebooted(const struct sockaddr_in *, const char *, unsigned int, u32); void nsm_release(struct nsm_handle *); +/* + * Host monitoring + */ +int nsm_monitor(const struct nlm_host *host); /* * This is used in garbage collection and resource reclaim diff --git a/include/linux/lockd/sm_inter.h b/include/linux/lockd/sm_inter.h index 5a5448bdb17..546b6102b0d 100644 --- a/include/linux/lockd/sm_inter.h +++ b/include/linux/lockd/sm_inter.h @@ -41,7 +41,6 @@ struct nsm_res { u32 state; }; -int nsm_monitor(struct nlm_host *); int nsm_unmonitor(struct nlm_host *); extern int nsm_local_state; -- cgit v1.2.3-70-g09d2 From c8c23c423dec49cb439697d3dc714e1500ff1610 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:31 -0500 Subject: NSM: Release nsmhandle in nlm_destroy_host The nsm_handle's reference count is bumped in nlm_lookup_host(). It should be decremented in nlm_destroy_host() to make it easier to see the balance of these two operations. Move the nsm_release() call to fs/lockd/host.c. The h_nsmhandle pointer is set in nlm_lookup_host(), and never cleared. The nlm_destroy_host() function is never called for the same nlm_host twice, so h_nsmhandle won't ever be NULL when nsm_unmonitor() is called. All references to the nlm_host are gone before it is freed. We can skip making h_nsmhandle NULL just before the nlm_host is deallocated. It's also likely we can remove the h_nsmhandle NULL check in nlmsvc_is_client() as well, but we can do that later when rearchitect- ing the nlm_host cache. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 8 +++----- fs/lockd/mon.c | 5 ----- include/linux/lockd/lockd.h | 1 - 3 files changed, 3 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 780918acd6f..1d523c1a7b6 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -37,6 +37,7 @@ static struct nsm_handle *nsm_find(const struct sockaddr *sap, const char *hostname, const size_t hostname_len, const int create); +static void nsm_release(struct nsm_handle *nsm); struct nlm_lookup_host_info { const int server; /* search for server|client */ @@ -263,10 +264,8 @@ nlm_destroy_host(struct nlm_host *host) BUG_ON(!list_empty(&host->h_lockowners)); BUG_ON(atomic_read(&host->h_count)); - /* - * Release NSM handle and unmonitor host. - */ nsm_unmonitor(host); + nsm_release(host->h_nsmhandle); clnt = host->h_rpcclnt; if (clnt != NULL) @@ -711,8 +710,7 @@ found: /* * Release an NSM handle */ -void -nsm_release(struct nsm_handle *nsm) +static void nsm_release(struct nsm_handle *nsm) { if (!nsm) return; diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index aaaa08e7ae7..15fab22db02 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -117,10 +117,6 @@ nsm_unmonitor(struct nlm_host *host) struct nsm_res res; int status = 0; - if (nsm == NULL) - return 0; - host->h_nsmhandle = NULL; - if (atomic_read(&nsm->sm_count) == 1 && nsm->sm_monitored && !nsm->sm_sticky) { dprintk("lockd: nsm_unmonitor(%s)\n", nsm->sm_name); @@ -132,7 +128,6 @@ nsm_unmonitor(struct nlm_host *host) else nsm->sm_monitored = 0; } - nsm_release(nsm); return status; } diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index f15a4f5ccbf..30a6a9c1ce4 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -240,7 +240,6 @@ void nlm_release_host(struct nlm_host *); void nlm_shutdown_hosts(void); extern void nlm_host_rebooted(const struct sockaddr_in *, const char *, unsigned int, u32); -void nsm_release(struct nsm_handle *); /* * Host monitoring -- cgit v1.2.3-70-g09d2 From 356c3eb466fd1a12afd6448d90fba3922836e5f1 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:38 -0500 Subject: NLM: Move the public declaration of nsm_unmonitor() to lockd.h Clean up. Make the nlm_host argument "const," and move the public declaration to lockd.h. Add a documenting comment. Bruce observed that nsm_unmonitor()'s only caller doesn't care about its return code, so make nsm_unmonitor() return void. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 15 +++++++++------ include/linux/lockd/lockd.h | 1 + include/linux/lockd/sm_inter.h | 1 - 3 files changed, 10 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 15fab22db02..d61cdc61cb5 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -107,15 +107,19 @@ int nsm_monitor(const struct nlm_host *host) return status; } -/* - * Cease to monitor remote host +/** + * nsm_unmonitor - Unregister peer notification + * @host: pointer to nlm_host of peer to stop monitoring + * + * If this peer is monitored, this function sends an upcall to + * tell the local rpc.statd not to send this peer a notification + * when we reboot. */ -int -nsm_unmonitor(struct nlm_host *host) +void nsm_unmonitor(const struct nlm_host *host) { struct nsm_handle *nsm = host->h_nsmhandle; struct nsm_res res; - int status = 0; + int status; if (atomic_read(&nsm->sm_count) == 1 && nsm->sm_monitored && !nsm->sm_sticky) { @@ -128,7 +132,6 @@ nsm_unmonitor(struct nlm_host *host) else nsm->sm_monitored = 0; } - return status; } /* diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 30a6a9c1ce4..38344bfb814 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -245,6 +245,7 @@ extern void nlm_host_rebooted(const struct sockaddr_in *, const char *, * Host monitoring */ int nsm_monitor(const struct nlm_host *host); +void nsm_unmonitor(const struct nlm_host *host); /* * This is used in garbage collection and resource reclaim diff --git a/include/linux/lockd/sm_inter.h b/include/linux/lockd/sm_inter.h index 546b6102b0d..896a5e30332 100644 --- a/include/linux/lockd/sm_inter.h +++ b/include/linux/lockd/sm_inter.h @@ -41,7 +41,6 @@ struct nsm_res { u32 state; }; -int nsm_unmonitor(struct nlm_host *); extern int nsm_local_state; #endif /* LINUX_LOCKD_SM_INTER_H */ -- cgit v1.2.3-70-g09d2 From 0c7aef4569f8680951b7dee01dddffb9d2f809ff Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 4 Dec 2008 14:21:46 -0500 Subject: NSM: Check result of SM_UNMON upcall Make sure any error returned by rpc.statd during an SM_UNMON call is reported rather than ignored completely. There isn't much to do with such an error, but we should log it in any case. Similar to a recent change to nsm_monitor(). Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index d61cdc61cb5..3bb71e1b1e1 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -126,6 +126,8 @@ void nsm_unmonitor(const struct nlm_host *host) dprintk("lockd: nsm_unmonitor(%s)\n", nsm->sm_name); status = nsm_mon_unmon(nsm, SM_UNMON, &res); + if (res.status != 0) + status = -EIO; if (status < 0) printk(KERN_NOTICE "lockd: cannot unmonitor %s\n", nsm->sm_name); -- cgit v1.2.3-70-g09d2 From 9c1bfd037f7ff8badaecb47418f109148d88bf45 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:01:59 -0500 Subject: NSM: Move NSM-related XDR data structures to lockd's xdr.h Clean up: NSM's XDR data structures are used only in fs/lockd/mon.c, so move them there. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 14 ++++++++++++++ include/linux/lockd/sm_inter.h | 20 -------------------- 2 files changed, 14 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 3bb71e1b1e1..81308832e99 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -18,6 +18,20 @@ #define NLMDBG_FACILITY NLMDBG_MONITOR +struct nsm_args { + __be32 addr; /* remote address */ + u32 prog; /* RPC callback info */ + u32 vers; + u32 proc; + + char *mon_name; +}; + +struct nsm_res { + u32 status; + u32 state; +}; + static struct rpc_clnt * nsm_create(void); static struct rpc_program nsm_program; diff --git a/include/linux/lockd/sm_inter.h b/include/linux/lockd/sm_inter.h index 896a5e30332..dd9d8a5bb31 100644 --- a/include/linux/lockd/sm_inter.h +++ b/include/linux/lockd/sm_inter.h @@ -21,26 +21,6 @@ #define SM_MAXSTRLEN 1024 #define SM_PRIV_SIZE 16 -/* - * Arguments for all calls to statd - */ -struct nsm_args { - __be32 addr; /* remote address */ - u32 prog; /* RPC callback info */ - u32 vers; - u32 proc; - - char * mon_name; -}; - -/* - * Result returned by statd - */ -struct nsm_res { - u32 status; - u32 state; -}; - extern int nsm_local_state; #endif /* LINUX_LOCKD_SM_INTER_H */ -- cgit v1.2.3-70-g09d2 From 36e8e668d3e6a61848a8921ddeb663b417299fa5 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:02:07 -0500 Subject: NSM: Move NSM program and procedure numbers to fs/lockd/mon.c Clean up: Move the RPC program and procedure numbers for NSM into the one source file that needs them: fs/lockd/mon.c. And, as with NLM, NFS, and rpcbind calls, use NSMPROC_FOO instead of SM_FOO for NSM procedure numbers. Finally, make a couple of comments more precise: what is referred to here as SM_NOTIFY is really the NLM (lockd) NLMPROC_SM_NOTIFY downcall, not NSMPROC_NOTIFY. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 42 +++++++++++++++++++++++++++--------------- include/linux/lockd/sm_inter.h | 9 --------- 2 files changed, 27 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 81308832e99..0fc9836db4e 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -17,6 +17,18 @@ #define NLMDBG_FACILITY NLMDBG_MONITOR +#define NSM_PROGRAM 100024 +#define NSM_VERSION 1 + +enum { + NSMPROC_NULL, + NSMPROC_STAT, + NSMPROC_MON, + NSMPROC_UNMON, + NSMPROC_UNMON_ALL, + NSMPROC_SIMU_CRASH, + NSMPROC_NOTIFY, +}; struct nsm_args { __be32 addr; /* remote address */ @@ -42,7 +54,7 @@ static struct rpc_program nsm_program; int nsm_local_state; /* - * Common procedure for SM_MON/SM_UNMON calls + * Common procedure for NSMPROC_MON/NSMPROC_UNMON calls */ static int nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) @@ -111,7 +123,7 @@ int nsm_monitor(const struct nlm_host *host) */ nsm->sm_mon_name = nsm_use_hostnames ? nsm->sm_name : nsm->sm_addrbuf; - status = nsm_mon_unmon(nsm, SM_MON, &res); + status = nsm_mon_unmon(nsm, NSMPROC_MON, &res); if (res.status != 0) status = -EIO; if (status < 0) @@ -139,7 +151,7 @@ void nsm_unmonitor(const struct nlm_host *host) && nsm->sm_monitored && !nsm->sm_sticky) { dprintk("lockd: nsm_unmonitor(%s)\n", nsm->sm_name); - status = nsm_mon_unmon(nsm, SM_UNMON, &res); + status = nsm_mon_unmon(nsm, NSMPROC_UNMON, &res); if (res.status != 0) status = -EIO; if (status < 0) @@ -167,7 +179,7 @@ nsm_create(void) .addrsize = sizeof(sin), .servername = "localhost", .program = &nsm_program, - .version = SM_VERSION, + .version = NSM_VERSION, .authflavor = RPC_AUTH_NULL, }; @@ -201,7 +213,7 @@ static __be32 *xdr_encode_mon_name(__be32 *p, struct nsm_args *argp) /* * The "my_id" argument specifies the hostname and RPC procedure * to be called when the status manager receives notification - * (via the SM_NOTIFY call) that the state of host "mon_name" + * (via the NLMPROC_SM_NOTIFY call) that the state of host "mon_name" * has changed. */ static __be32 *xdr_encode_my_id(__be32 *p, struct nsm_args *argp) @@ -219,7 +231,7 @@ static __be32 *xdr_encode_my_id(__be32 *p, struct nsm_args *argp) /* * The "mon_id" argument specifies the non-private arguments - * of an SM_MON or SM_UNMON call. + * of an NSMPROC_MON or NSMPROC_UNMON call. */ static __be32 *xdr_encode_mon_id(__be32 *p, struct nsm_args *argp) { @@ -232,8 +244,8 @@ static __be32 *xdr_encode_mon_id(__be32 *p, struct nsm_args *argp) /* * The "priv" argument may contain private information required - * by the SM_MON call. This information will be supplied in the - * SM_NOTIFY call. + * by the NSMPROC_MON call. This information will be supplied in the + * NLMPROC_SM_NOTIFY call. * * Linux provides the raw IP address of the monitored host, * left in network byte order. @@ -300,22 +312,22 @@ xdr_decode_stat(struct rpc_rqst *rqstp, __be32 *p, struct nsm_res *resp) #define SM_unmonres_sz 1 static struct rpc_procinfo nsm_procedures[] = { -[SM_MON] = { - .p_proc = SM_MON, +[NSMPROC_MON] = { + .p_proc = NSMPROC_MON, .p_encode = (kxdrproc_t) xdr_encode_mon, .p_decode = (kxdrproc_t) xdr_decode_stat_res, .p_arglen = SM_mon_sz, .p_replen = SM_monres_sz, - .p_statidx = SM_MON, + .p_statidx = NSMPROC_MON, .p_name = "MONITOR", }, -[SM_UNMON] = { - .p_proc = SM_UNMON, +[NSMPROC_UNMON] = { + .p_proc = NSMPROC_UNMON, .p_encode = (kxdrproc_t) xdr_encode_unmon, .p_decode = (kxdrproc_t) xdr_decode_stat, .p_arglen = SM_mon_id_sz, .p_replen = SM_unmonres_sz, - .p_statidx = SM_UNMON, + .p_statidx = NSMPROC_UNMON, .p_name = "UNMONITOR", }, }; @@ -334,7 +346,7 @@ static struct rpc_stat nsm_stats; static struct rpc_program nsm_program = { .name = "statd", - .number = SM_PROGRAM, + .number = NSM_PROGRAM, .nrvers = ARRAY_SIZE(nsm_version), .version = nsm_version, .stats = &nsm_stats diff --git a/include/linux/lockd/sm_inter.h b/include/linux/lockd/sm_inter.h index dd9d8a5bb31..116bf38535a 100644 --- a/include/linux/lockd/sm_inter.h +++ b/include/linux/lockd/sm_inter.h @@ -9,15 +9,6 @@ #ifndef LINUX_LOCKD_SM_INTER_H #define LINUX_LOCKD_SM_INTER_H -#define SM_PROGRAM 100024 -#define SM_VERSION 1 -#define SM_STAT 1 -#define SM_MON 2 -#define SM_UNMON 3 -#define SM_UNMON_ALL 4 -#define SM_SIMU_CRASH 5 -#define SM_NOTIFY 6 - #define SM_MAXSTRLEN 1024 #define SM_PRIV_SIZE 16 -- cgit v1.2.3-70-g09d2 From 03eb1dcbb799304b58730f4dba65812f49fb305e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:02:15 -0500 Subject: NSM: move to xdr_stream-based XDR encoders and decoders Introduce xdr_stream-based XDR encoder and decoder functions, which are more careful about preventing RPC buffer overflows. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 130 ++++++++++++++++++++++++++++++++++----------------------- 1 file changed, 78 insertions(+), 52 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 0fc9836db4e..81e1cc14246 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -193,21 +193,26 @@ nsm_create(void) * Status Monitor wire protocol. */ -static __be32 *xdr_encode_nsm_string(__be32 *p, char *string) +static int encode_nsm_string(struct xdr_stream *xdr, const char *string) { - size_t len = strlen(string); - - if (len > SM_MAXSTRLEN) - len = SM_MAXSTRLEN; - return xdr_encode_opaque(p, string, len); + const u32 len = strlen(string); + __be32 *p; + + if (unlikely(len > SM_MAXSTRLEN)) + return -EIO; + p = xdr_reserve_space(xdr, sizeof(u32) + len); + if (unlikely(p == NULL)) + return -EIO; + xdr_encode_opaque(p, string, len); + return 0; } /* * "mon_name" specifies the host to be monitored. */ -static __be32 *xdr_encode_mon_name(__be32 *p, struct nsm_args *argp) +static int encode_mon_name(struct xdr_stream *xdr, const struct nsm_args *argp) { - return xdr_encode_nsm_string(p, argp->mon_name); + return encode_nsm_string(xdr, argp->mon_name); } /* @@ -216,30 +221,35 @@ static __be32 *xdr_encode_mon_name(__be32 *p, struct nsm_args *argp) * (via the NLMPROC_SM_NOTIFY call) that the state of host "mon_name" * has changed. */ -static __be32 *xdr_encode_my_id(__be32 *p, struct nsm_args *argp) +static int encode_my_id(struct xdr_stream *xdr, const struct nsm_args *argp) { - p = xdr_encode_nsm_string(p, utsname()->nodename); - if (!p) - return ERR_PTR(-EIO); - + int status; + __be32 *p; + + status = encode_nsm_string(xdr, utsname()->nodename); + if (unlikely(status != 0)) + return status; + p = xdr_reserve_space(xdr, 3 * sizeof(u32)); + if (unlikely(p == NULL)) + return -EIO; *p++ = htonl(argp->prog); *p++ = htonl(argp->vers); *p++ = htonl(argp->proc); - - return p; + return 0; } /* * The "mon_id" argument specifies the non-private arguments * of an NSMPROC_MON or NSMPROC_UNMON call. */ -static __be32 *xdr_encode_mon_id(__be32 *p, struct nsm_args *argp) +static int encode_mon_id(struct xdr_stream *xdr, const struct nsm_args *argp) { - p = xdr_encode_mon_name(p, argp); - if (!p) - return ERR_PTR(-EIO); + int status; - return xdr_encode_my_id(p, argp); + status = encode_mon_name(xdr, argp); + if (unlikely(status != 0)) + return status; + return encode_my_id(xdr, argp); } /* @@ -250,55 +260,71 @@ static __be32 *xdr_encode_mon_id(__be32 *p, struct nsm_args *argp) * Linux provides the raw IP address of the monitored host, * left in network byte order. */ -static __be32 *xdr_encode_priv(__be32 *p, struct nsm_args *argp) +static int encode_priv(struct xdr_stream *xdr, const struct nsm_args *argp) { + __be32 *p; + + p = xdr_reserve_space(xdr, SM_PRIV_SIZE); + if (unlikely(p == NULL)) + return -EIO; *p++ = argp->addr; *p++ = 0; *p++ = 0; *p++ = 0; - - return p; + return 0; } -static int -xdr_encode_mon(struct rpc_rqst *rqstp, __be32 *p, struct nsm_args *argp) +static int xdr_enc_mon(struct rpc_rqst *req, __be32 *p, + const struct nsm_args *argp) { - p = xdr_encode_mon_id(p, argp); - if (IS_ERR(p)) - return PTR_ERR(p); - - p = xdr_encode_priv(p, argp); - if (IS_ERR(p)) - return PTR_ERR(p); + struct xdr_stream xdr; + int status; - rqstp->rq_slen = xdr_adjust_iovec(rqstp->rq_svec, p); - return 0; + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + status = encode_mon_id(&xdr, argp); + if (unlikely(status)) + return status; + return encode_priv(&xdr, argp); } -static int -xdr_encode_unmon(struct rpc_rqst *rqstp, __be32 *p, struct nsm_args *argp) +static int xdr_enc_unmon(struct rpc_rqst *req, __be32 *p, + const struct nsm_args *argp) { - p = xdr_encode_mon_id(p, argp); - if (IS_ERR(p)) - return PTR_ERR(p); - rqstp->rq_slen = xdr_adjust_iovec(rqstp->rq_svec, p); - return 0; + struct xdr_stream xdr; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + return encode_mon_id(&xdr, argp); } -static int -xdr_decode_stat_res(struct rpc_rqst *rqstp, __be32 *p, struct nsm_res *resp) +static int xdr_dec_stat_res(struct rpc_rqst *rqstp, __be32 *p, + struct nsm_res *resp) { + struct xdr_stream xdr; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + p = xdr_inline_decode(&xdr, 2 * sizeof(u32)); + if (unlikely(p == NULL)) + return -EIO; resp->status = ntohl(*p++); - resp->state = ntohl(*p++); - dprintk("nsm: xdr_decode_stat_res status %d state %d\n", + resp->state = ntohl(*p); + + dprintk("lockd: xdr_dec_stat_res status %d state %d\n", resp->status, resp->state); return 0; } -static int -xdr_decode_stat(struct rpc_rqst *rqstp, __be32 *p, struct nsm_res *resp) +static int xdr_dec_stat(struct rpc_rqst *rqstp, __be32 *p, + struct nsm_res *resp) { - resp->state = ntohl(*p++); + struct xdr_stream xdr; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + p = xdr_inline_decode(&xdr, sizeof(u32)); + if (unlikely(p == NULL)) + return -EIO; + resp->state = ntohl(*p); + + dprintk("lockd: xdr_dec_stat state %d\n", resp->state); return 0; } @@ -314,8 +340,8 @@ xdr_decode_stat(struct rpc_rqst *rqstp, __be32 *p, struct nsm_res *resp) static struct rpc_procinfo nsm_procedures[] = { [NSMPROC_MON] = { .p_proc = NSMPROC_MON, - .p_encode = (kxdrproc_t) xdr_encode_mon, - .p_decode = (kxdrproc_t) xdr_decode_stat_res, + .p_encode = (kxdrproc_t)xdr_enc_mon, + .p_decode = (kxdrproc_t)xdr_dec_stat_res, .p_arglen = SM_mon_sz, .p_replen = SM_monres_sz, .p_statidx = NSMPROC_MON, @@ -323,8 +349,8 @@ static struct rpc_procinfo nsm_procedures[] = { }, [NSMPROC_UNMON] = { .p_proc = NSMPROC_UNMON, - .p_encode = (kxdrproc_t) xdr_encode_unmon, - .p_decode = (kxdrproc_t) xdr_decode_stat, + .p_encode = (kxdrproc_t)xdr_enc_unmon, + .p_decode = (kxdrproc_t)xdr_dec_stat, .p_arglen = SM_mon_id_sz, .p_replen = SM_unmonres_sz, .p_statidx = NSMPROC_UNMON, -- cgit v1.2.3-70-g09d2 From 67c6d107a689243979a2b5f15244b5261634a924 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:02:45 -0500 Subject: NSM: Move nsm_find() to fs/lockd/mon.c The nsm_find() function sets up fresh nsm_handle entries. This is where we will store the "priv" cookie used to lookup nsm_handles during reboot recovery. The cookie will be constructed when nsm_find() creates a new nsm_handle. As much as possible, I would like to keep everything that handles a "priv" cookie in fs/lockd/mon.c so that all the smarts are in one source file. That organization should make it pretty simple to see how all this works. To me, it makes more sense than the current arrangement to keep nsm_find() with nsm_monitor() and nsm_unmonitor(). So, start reorganizing by moving nsm_find() into fs/lockd/mon.c. The nsm_release() function comes along too, since it shares the nsm_lock global variable. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 128 ------------------------------------------ fs/lockd/mon.c | 133 ++++++++++++++++++++++++++++++++++++++++++++ include/linux/lockd/lockd.h | 6 ++ 3 files changed, 139 insertions(+), 128 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 1d523c1a7b6..dbdeaa88d2f 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -32,12 +32,6 @@ static int nrhosts; static DEFINE_MUTEX(nlm_host_mutex); static void nlm_gc_hosts(void); -static struct nsm_handle *nsm_find(const struct sockaddr *sap, - const size_t salen, - const char *hostname, - const size_t hostname_len, - const int create); -static void nsm_release(struct nsm_handle *nsm); struct nlm_lookup_host_info { const int server; /* search for server|client */ @@ -106,43 +100,6 @@ static void nlm_clear_port(struct sockaddr *sap) } } -static void nlm_display_ipv4_address(const struct sockaddr *sap, char *buf, - const size_t len) -{ - const struct sockaddr_in *sin = (struct sockaddr_in *)sap; - snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); -} - -static void nlm_display_ipv6_address(const struct sockaddr *sap, char *buf, - const size_t len) -{ - const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; - - if (ipv6_addr_v4mapped(&sin6->sin6_addr)) - snprintf(buf, len, "%pI4", &sin6->sin6_addr.s6_addr32[3]); - else if (sin6->sin6_scope_id != 0) - snprintf(buf, len, "%pI6%%%u", &sin6->sin6_addr, - sin6->sin6_scope_id); - else - snprintf(buf, len, "%pI6", &sin6->sin6_addr); -} - -static void nlm_display_address(const struct sockaddr *sap, - char *buf, const size_t len) -{ - switch (sap->sa_family) { - case AF_INET: - nlm_display_ipv4_address(sap, buf, len); - break; - case AF_INET6: - nlm_display_ipv6_address(sap, buf, len); - break; - default: - snprintf(buf, len, "unsupported address family"); - break; - } -} - /* * Common host lookup routine for server & client */ @@ -635,88 +592,3 @@ nlm_gc_hosts(void) next_gc = jiffies + NLM_HOST_COLLECT; } - - -/* - * Manage NSM handles - */ -static LIST_HEAD(nsm_handles); -static DEFINE_SPINLOCK(nsm_lock); - -static struct nsm_handle *nsm_find(const struct sockaddr *sap, - const size_t salen, - const char *hostname, - const size_t hostname_len, - const int create) -{ - struct nsm_handle *nsm = NULL; - struct nsm_handle *pos; - - if (!sap) - return NULL; - - if (hostname && memchr(hostname, '/', hostname_len) != NULL) { - if (printk_ratelimit()) { - printk(KERN_WARNING "Invalid hostname \"%.*s\" " - "in NFS lock request\n", - (int)hostname_len, hostname); - } - return NULL; - } - -retry: - spin_lock(&nsm_lock); - list_for_each_entry(pos, &nsm_handles, sm_link) { - - if (hostname && nsm_use_hostnames) { - if (strlen(pos->sm_name) != hostname_len - || memcmp(pos->sm_name, hostname, hostname_len)) - continue; - } else if (!nlm_cmp_addr(nsm_addr(pos), sap)) - continue; - atomic_inc(&pos->sm_count); - kfree(nsm); - nsm = pos; - goto found; - } - if (nsm) { - list_add(&nsm->sm_link, &nsm_handles); - goto found; - } - spin_unlock(&nsm_lock); - - if (!create) - return NULL; - - nsm = kzalloc(sizeof(*nsm) + hostname_len + 1, GFP_KERNEL); - if (nsm == NULL) - return NULL; - - memcpy(nsm_addr(nsm), sap, salen); - nsm->sm_addrlen = salen; - nsm->sm_name = (char *) (nsm + 1); - memcpy(nsm->sm_name, hostname, hostname_len); - nsm->sm_name[hostname_len] = '\0'; - nlm_display_address((struct sockaddr *)&nsm->sm_addr, - nsm->sm_addrbuf, sizeof(nsm->sm_addrbuf)); - atomic_set(&nsm->sm_count, 1); - goto retry; - -found: - spin_unlock(&nsm_lock); - return nsm; -} - -/* - * Release an NSM handle - */ -static void nsm_release(struct nsm_handle *nsm) -{ - if (!nsm) - return; - if (atomic_dec_and_lock(&nsm->sm_count, &nsm_lock)) { - list_del(&nsm->sm_link); - spin_unlock(&nsm_lock); - kfree(nsm); - } -} diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 81e1cc14246..8e68e799293 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -47,12 +47,51 @@ struct nsm_res { static struct rpc_clnt * nsm_create(void); static struct rpc_program nsm_program; +static LIST_HEAD(nsm_handles); +static DEFINE_SPINLOCK(nsm_lock); /* * Local NSM state */ int nsm_local_state; +static void nsm_display_ipv4_address(const struct sockaddr *sap, char *buf, + const size_t len) +{ + const struct sockaddr_in *sin = (struct sockaddr_in *)sap; + snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); +} + +static void nsm_display_ipv6_address(const struct sockaddr *sap, char *buf, + const size_t len) +{ + const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; + + if (ipv6_addr_v4mapped(&sin6->sin6_addr)) + snprintf(buf, len, "%pI4", &sin6->sin6_addr.s6_addr32[3]); + else if (sin6->sin6_scope_id != 0) + snprintf(buf, len, "%pI6%%%u", &sin6->sin6_addr, + sin6->sin6_scope_id); + else + snprintf(buf, len, "%pI6", &sin6->sin6_addr); +} + +static void nsm_display_address(const struct sockaddr *sap, + char *buf, const size_t len) +{ + switch (sap->sa_family) { + case AF_INET: + nsm_display_ipv4_address(sap, buf, len); + break; + case AF_INET6: + nsm_display_ipv6_address(sap, buf, len); + break; + default: + snprintf(buf, len, "unsupported address family"); + break; + } +} + /* * Common procedure for NSMPROC_MON/NSMPROC_UNMON calls */ @@ -162,6 +201,100 @@ void nsm_unmonitor(const struct nlm_host *host) } } +/** + * nsm_find - Find or create a cached nsm_handle + * @sap: pointer to socket address of handle to find + * @salen: length of socket address + * @hostname: pointer to C string containing hostname to find + * @hostname_len: length of C string + * @create: one means create new handle if not found in cache + * + * Behavior is modulated by the global nsm_use_hostnames variable + * and by the @create argument. + * + * Returns a cached nsm_handle after bumping its ref count, or if + * @create is set, returns a fresh nsm_handle if a handle that + * matches @sap and/or @hostname cannot be found in the handle cache. + * Returns NULL if an error occurs. + */ +struct nsm_handle *nsm_find(const struct sockaddr *sap, const size_t salen, + const char *hostname, const size_t hostname_len, + const int create) +{ + struct nsm_handle *nsm = NULL; + struct nsm_handle *pos; + + if (!sap) + return NULL; + + if (hostname && memchr(hostname, '/', hostname_len) != NULL) { + if (printk_ratelimit()) { + printk(KERN_WARNING "Invalid hostname \"%.*s\" " + "in NFS lock request\n", + (int)hostname_len, hostname); + } + return NULL; + } + +retry: + spin_lock(&nsm_lock); + list_for_each_entry(pos, &nsm_handles, sm_link) { + + if (hostname && nsm_use_hostnames) { + if (strlen(pos->sm_name) != hostname_len + || memcmp(pos->sm_name, hostname, hostname_len)) + continue; + } else if (!nlm_cmp_addr(nsm_addr(pos), sap)) + continue; + atomic_inc(&pos->sm_count); + kfree(nsm); + nsm = pos; + goto found; + } + if (nsm) { + list_add(&nsm->sm_link, &nsm_handles); + goto found; + } + spin_unlock(&nsm_lock); + + if (!create) + return NULL; + + nsm = kzalloc(sizeof(*nsm) + hostname_len + 1, GFP_KERNEL); + if (nsm == NULL) + return NULL; + + memcpy(nsm_addr(nsm), sap, salen); + nsm->sm_addrlen = salen; + nsm->sm_name = (char *) (nsm + 1); + memcpy(nsm->sm_name, hostname, hostname_len); + nsm->sm_name[hostname_len] = '\0'; + nsm_display_address((struct sockaddr *)&nsm->sm_addr, + nsm->sm_addrbuf, sizeof(nsm->sm_addrbuf)); + atomic_set(&nsm->sm_count, 1); + goto retry; + +found: + spin_unlock(&nsm_lock); + return nsm; +} + +/** + * nsm_release - Release an NSM handle + * @nsm: pointer to handle to be released + * + */ +void nsm_release(struct nsm_handle *nsm) +{ + if (!nsm) + return; + if (atomic_dec_and_lock(&nsm->sm_count, &nsm_lock)) { + list_del(&nsm->sm_link); + spin_unlock(&nsm_lock); + kfree(nsm); + } +} + /* * Create NSM client for the local host */ diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 38344bfb814..8d715363c6a 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -247,6 +247,12 @@ extern void nlm_host_rebooted(const struct sockaddr_in *, const char *, int nsm_monitor(const struct nlm_host *host); void nsm_unmonitor(const struct nlm_host *host); +struct nsm_handle *nsm_find(const struct sockaddr *sap, const size_t salen, + const char *hostname, + const size_t hostname_len, + const int create); +void nsm_release(struct nsm_handle *nsm); + /* * This is used in garbage collection and resource reclaim * A return value != 0 means destroy the lock/block/share -- cgit v1.2.3-70-g09d2 From 5cf1c4b19db99d21d44c2ab457cfd44eb86b4439 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:02:53 -0500 Subject: NSM: Add dprintk() calls in nsm_find and nsm_release Introduce some dprintk() calls in fs/lockd/mon.c that are enabled by the NLMDBG_MONITOR flag. These report when we find, create, and release nsm_handles. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 8e68e799293..38255455563 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -249,10 +249,15 @@ retry: atomic_inc(&pos->sm_count); kfree(nsm); nsm = pos; + dprintk("lockd: found nsm_handle for %s (%s), cnt %d\n", + pos->sm_name, pos->sm_addrbuf, + atomic_read(&pos->sm_count)); goto found; } if (nsm) { list_add(&nsm->sm_link, &nsm_handles); + dprintk("lockd: created nsm_handle for %s (%s)\n", + nsm->sm_name, nsm->sm_addrbuf); goto found; } spin_unlock(&nsm_lock); @@ -291,6 +296,8 @@ void nsm_release(struct nsm_handle *nsm) if (atomic_dec_and_lock(&nsm->sm_count, &nsm_lock)) { list_del(&nsm->sm_link); spin_unlock(&nsm_lock); + dprintk("lockd: destroyed nsm_handle for %s (%s)\n", + nsm->sm_name, nsm->sm_addrbuf); kfree(nsm); } } -- cgit v1.2.3-70-g09d2 From bc1cc6c4e476b60df48227165990c87a22db6bb7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:01 -0500 Subject: NSM: Remove NULL pointer check from nsm_find() The nsm_find() function should never be called with a NULL IP address pointer. If it is, that's a bug. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 38255455563..0a066a13478 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -224,9 +224,6 @@ struct nsm_handle *nsm_find(const struct sockaddr *sap, const size_t salen, struct nsm_handle *nsm = NULL; struct nsm_handle *pos; - if (!sap) - return NULL; - if (hostname && memchr(hostname, '/', hostname_len) != NULL) { if (printk_ratelimit()) { printk(KERN_WARNING "Invalid hostname \"%.*s\" " -- cgit v1.2.3-70-g09d2 From 05f3a9af58180d24a9decedd71d4587935782d70 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:09 -0500 Subject: NSM: Remove !nsm check from nsm_release() The nsm_release() function should never be called with a NULL handle point. If it is, that's a bug. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 0a066a13478..0792900b628 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -288,8 +288,6 @@ found: */ void nsm_release(struct nsm_handle *nsm) { - if (!nsm) - return; if (atomic_dec_and_lock(&nsm->sm_count, &nsm_lock)) { list_del(&nsm->sm_link); spin_unlock(&nsm_lock); -- cgit v1.2.3-70-g09d2 From 7e44d3bea21fbb9494930d1cd35ca92a9a4a3279 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:16 -0500 Subject: NSM: Generate NSMPROC_MON's "priv" argument when nsm_handle is created Introduce a new data type, used by both the in-kernel NLM and NSM implementations, that is used to manage the opaque "priv" argument for the NSMPROC_MON and NLMPROC_SM_NOTIFY calls. Construct the "priv" cookie when the nsm_handle is created. The nsm_init_private() function may look a little strange, but it is roughly equivalent to how the XDR encoder formed the "priv" argument. It's going to go away soon. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 16 ++++++++++++++++ include/linux/lockd/lockd.h | 1 + include/linux/lockd/sm_inter.h | 1 - include/linux/lockd/xdr.h | 6 ++++++ 4 files changed, 23 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 0792900b628..c8d18cd22b8 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -201,6 +201,21 @@ void nsm_unmonitor(const struct nlm_host *host) } } +/* + * Construct a unique cookie to match this nsm_handle to this monitored + * host. It is passed to the local rpc.statd via NSMPROC_MON, and + * returned via NLMPROC_SM_NOTIFY, in the "priv" field of these + * requests. + * + * Linux provides the raw IP address of the monitored host, + * left in network byte order. + */ +static void nsm_init_private(struct nsm_handle *nsm) +{ + __be32 *p = (__be32 *)&nsm->sm_priv.data; + *p = nsm_addr_in(nsm)->sin_addr.s_addr; +} + /** * nsm_find - Find or create a cached nsm_handle * @sap: pointer to socket address of handle to find @@ -271,6 +286,7 @@ retry: nsm->sm_name = (char *) (nsm + 1); memcpy(nsm->sm_name, hostname, hostname_len); nsm->sm_name[hostname_len] = '\0'; + nsm_init_private(nsm); nsm_display_address((struct sockaddr *)&nsm->sm_addr, nsm->sm_addrbuf, sizeof(nsm->sm_addrbuf)); atomic_set(&nsm->sm_count, 1); diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 8d715363c6a..194fa8a6639 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -85,6 +85,7 @@ struct nsm_handle { size_t sm_addrlen; unsigned int sm_monitored : 1, sm_sticky : 1; /* don't unmonitor */ + struct nsm_private sm_priv; char sm_addrbuf[NSM_ADDRBUF]; }; diff --git a/include/linux/lockd/sm_inter.h b/include/linux/lockd/sm_inter.h index 116bf38535a..5cef5a79dd9 100644 --- a/include/linux/lockd/sm_inter.h +++ b/include/linux/lockd/sm_inter.h @@ -10,7 +10,6 @@ #define LINUX_LOCKD_SM_INTER_H #define SM_MAXSTRLEN 1024 -#define SM_PRIV_SIZE 16 extern int nsm_local_state; diff --git a/include/linux/lockd/xdr.h b/include/linux/lockd/xdr.h index d6b3a802c04..6b519926385 100644 --- a/include/linux/lockd/xdr.h +++ b/include/linux/lockd/xdr.h @@ -13,6 +13,12 @@ #include #include +#define SM_PRIV_SIZE 16 + +struct nsm_private { + unsigned char data[SM_PRIV_SIZE]; +}; + struct svc_rqst; #define NLM_MAXCOOKIELEN 32 -- cgit v1.2.3-70-g09d2 From cab2d3c99165abbba2943f1b269003b17fd3b1cb Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:24 -0500 Subject: NSM: Encode the new "priv" cookie for NSMPROC_MON requests Pass the new "priv" cookie to NSMPROC_MON's XDR encoder, instead of creating the "priv" argument in the encoder at call time. This patch should not cause a behavioral change: the contents of the cookie remain the same for the time being. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index c8d18cd22b8..4424b0a5a51 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -31,7 +31,7 @@ enum { }; struct nsm_args { - __be32 addr; /* remote address */ + struct nsm_private *priv; u32 prog; /* RPC callback info */ u32 vers; u32 proc; @@ -101,7 +101,7 @@ nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) struct rpc_clnt *clnt; int status; struct nsm_args args = { - .addr = nsm_addr_in(nsm)->sin_addr.s_addr, + .priv = &nsm->sm_priv, .prog = NLM_PROGRAM, .vers = 3, .proc = NLMPROC_NSM_NOTIFY, @@ -407,9 +407,6 @@ static int encode_mon_id(struct xdr_stream *xdr, const struct nsm_args *argp) * The "priv" argument may contain private information required * by the NSMPROC_MON call. This information will be supplied in the * NLMPROC_SM_NOTIFY call. - * - * Linux provides the raw IP address of the monitored host, - * left in network byte order. */ static int encode_priv(struct xdr_stream *xdr, const struct nsm_args *argp) { @@ -418,10 +415,7 @@ static int encode_priv(struct xdr_stream *xdr, const struct nsm_args *argp) p = xdr_reserve_space(xdr, SM_PRIV_SIZE); if (unlikely(p == NULL)) return -EIO; - *p++ = argp->addr; - *p++ = 0; - *p++ = 0; - *p++ = 0; + xdr_encode_opaque_fixed(p, argp->priv->data, SM_PRIV_SIZE); return 0; } -- cgit v1.2.3-70-g09d2 From 7fefc9cb9d5f129c238d93166f705c96ca2e7e51 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:31 -0500 Subject: NLM: Change nlm_host_rebooted() to take a single nlm_reboot argument Pass the nlm_reboot data structure directly from the NLMPROC_SM_NOTIFY XDR decoders to nlm_host_rebooted(). This eliminates some packing and unpacking of the NLMPROC_SM_NOTIFY results, and prepares for passing these results, including the "priv" cookie, directly to a lookup routine in fs/lockd/mon.c. This patch changes code organization but should not cause any behavioral change. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 31 +++++++++++++++++-------------- fs/lockd/svc4proc.c | 11 +---------- fs/lockd/svcproc.c | 11 +---------- include/linux/lockd/lockd.h | 3 +-- 4 files changed, 20 insertions(+), 36 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index dbdeaa88d2f..ed103387964 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -444,31 +444,34 @@ void nlm_release_host(struct nlm_host *host) } } -/* - * We were notified that the host indicated by address &sin - * has rebooted. - * Release all resources held by that peer. +/** + * nlm_host_rebooted - Release all resources held by rebooted host + * @info: pointer to decoded results of NLM_SM_NOTIFY call + * + * We were notified that the specified host has rebooted. Release + * all resources held by that peer. */ -void nlm_host_rebooted(const struct sockaddr_in *sin, - const char *hostname, - unsigned int hostname_len, - u32 new_state) +void nlm_host_rebooted(const struct nlm_reboot *info) { + const struct sockaddr_in sin = { + .sin_family = AF_INET, + .sin_addr.s_addr = info->addr, + }; struct hlist_head *chain; struct hlist_node *pos; struct nsm_handle *nsm; struct nlm_host *host; - nsm = nsm_find((struct sockaddr *)sin, sizeof(*sin), - hostname, hostname_len, 0); + nsm = nsm_find((struct sockaddr *)&sin, sizeof(sin), + info->mon, info->len, 0); if (nsm == NULL) { dprintk("lockd: never saw rebooted peer '%.*s' before\n", - hostname_len, hostname); + info->len, info->mon); return; } dprintk("lockd: nlm_host_rebooted(%.*s, %s)\n", - hostname_len, hostname, nsm->sm_addrbuf); + info->len, info->mon, nsm->sm_addrbuf); /* When reclaiming locks on this peer, make sure that * we set up a new notification */ @@ -483,8 +486,8 @@ again: mutex_lock(&nlm_host_mutex); for (chain = nlm_hosts; chain < nlm_hosts + NLM_HOST_NRHASH; ++chain) { hlist_for_each_entry(host, pos, chain, h_hash) { if (host->h_nsmhandle == nsm - && host->h_nsmstate != new_state) { - host->h_nsmstate = new_state; + && host->h_nsmstate != info->state) { + host->h_nsmstate = info->state; host->h_state++; nlm_get_host(host); diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c index 4dfdcbc6bf6..bb79a53e060 100644 --- a/fs/lockd/svc4proc.c +++ b/fs/lockd/svc4proc.c @@ -419,8 +419,6 @@ static __be32 nlm4svc_proc_sm_notify(struct svc_rqst *rqstp, struct nlm_reboot *argp, void *resp) { - struct sockaddr_in saddr; - dprintk("lockd: SM_NOTIFY called\n"); if (!nlm_privileged_requester(rqstp)) { @@ -430,14 +428,7 @@ nlm4svc_proc_sm_notify(struct svc_rqst *rqstp, struct nlm_reboot *argp, return rpc_system_err; } - /* Obtain the host pointer for this NFS server and try to - * reclaim all locks we hold on this server. - */ - memset(&saddr, 0, sizeof(saddr)); - saddr.sin_family = AF_INET; - saddr.sin_addr.s_addr = argp->addr; - nlm_host_rebooted(&saddr, argp->mon, argp->len, argp->state); - + nlm_host_rebooted(argp); return rpc_success; } diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index 3ca89e2a938..e44310c0211 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -451,8 +451,6 @@ static __be32 nlmsvc_proc_sm_notify(struct svc_rqst *rqstp, struct nlm_reboot *argp, void *resp) { - struct sockaddr_in saddr; - dprintk("lockd: SM_NOTIFY called\n"); if (!nlm_privileged_requester(rqstp)) { @@ -462,14 +460,7 @@ nlmsvc_proc_sm_notify(struct svc_rqst *rqstp, struct nlm_reboot *argp, return rpc_system_err; } - /* Obtain the host pointer for this NFS server and try to - * reclaim all locks we hold on this server. - */ - memset(&saddr, 0, sizeof(saddr)); - saddr.sin_family = AF_INET; - saddr.sin_addr.s_addr = argp->addr; - nlm_host_rebooted(&saddr, argp->mon, argp->len, argp->state); - + nlm_host_rebooted(argp); return rpc_success; } diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 194fa8a6639..2a3533ea38d 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -239,8 +239,7 @@ void nlm_rebind_host(struct nlm_host *); struct nlm_host * nlm_get_host(struct nlm_host *); void nlm_release_host(struct nlm_host *); void nlm_shutdown_hosts(void); -extern void nlm_host_rebooted(const struct sockaddr_in *, const char *, - unsigned int, u32); +void nlm_host_rebooted(const struct nlm_reboot *); /* * Host monitoring -- cgit v1.2.3-70-g09d2 From 576df4634e37e46b441fefb91915184edb13bb94 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:39 -0500 Subject: NLM: Decode "priv" argument of NLMPROC_SM_NOTIFY as an opaque The NLM XDR decoders for the NLMPROC_SM_NOTIFY procedure should treat their "priv" argument truly as an opaque, as defined by the protocol, and let the upper layers figure out what is in it. This will make it easier to modify the contents and interpretation of the "priv" argument, and keep knowledge about what's in "priv" local to fs/lockd/mon.c. For now, the NLM and NSM implementations should behave exactly as they did before. The formation of the address of the rebooted host in nlm_host_rebooted() may look a little strange, but it is the inverse of how nsm_init_private() forms the private cookie. Plus, it's going away soon anyway. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 3 ++- fs/lockd/xdr.c | 4 ++-- fs/lockd/xdr4.c | 4 ++-- include/linux/lockd/xdr.h | 8 ++++---- 4 files changed, 10 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index ed103387964..dc41e46ef74 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -453,9 +453,10 @@ void nlm_release_host(struct nlm_host *host) */ void nlm_host_rebooted(const struct nlm_reboot *info) { + __be32 *p = (__be32 *)&info->priv.data; const struct sockaddr_in sin = { .sin_family = AF_INET, - .sin_addr.s_addr = info->addr, + .sin_addr.s_addr = *p, }; struct hlist_head *chain; struct hlist_node *pos; diff --git a/fs/lockd/xdr.c b/fs/lockd/xdr.c index 1f226290c67..4cc7d01a1eb 100644 --- a/fs/lockd/xdr.c +++ b/fs/lockd/xdr.c @@ -349,8 +349,8 @@ nlmsvc_decode_reboot(struct svc_rqst *rqstp, __be32 *p, struct nlm_reboot *argp) if (!(p = xdr_decode_string_inplace(p, &argp->mon, &argp->len, SM_MAXSTRLEN))) return 0; argp->state = ntohl(*p++); - /* Preserve the address in network byte order */ - argp->addr = *p++; + memcpy(&argp->priv.data, p, sizeof(argp->priv.data)); + p += XDR_QUADLEN(SM_PRIV_SIZE); return xdr_argsize_check(rqstp, p); } diff --git a/fs/lockd/xdr4.c b/fs/lockd/xdr4.c index 50c493a8ad8..61d1714a470 100644 --- a/fs/lockd/xdr4.c +++ b/fs/lockd/xdr4.c @@ -356,8 +356,8 @@ nlm4svc_decode_reboot(struct svc_rqst *rqstp, __be32 *p, struct nlm_reboot *argp if (!(p = xdr_decode_string_inplace(p, &argp->mon, &argp->len, SM_MAXSTRLEN))) return 0; argp->state = ntohl(*p++); - /* Preserve the address in network byte order */ - argp->addr = *p++; + memcpy(&argp->priv.data, p, sizeof(argp->priv.data)); + p += XDR_QUADLEN(SM_PRIV_SIZE); return xdr_argsize_check(rqstp, p); } diff --git a/include/linux/lockd/xdr.h b/include/linux/lockd/xdr.h index 6b519926385..6338866222a 100644 --- a/include/linux/lockd/xdr.h +++ b/include/linux/lockd/xdr.h @@ -83,10 +83,10 @@ struct nlm_res { * statd callback when client has rebooted */ struct nlm_reboot { - char * mon; - unsigned int len; - u32 state; - __be32 addr; + char *mon; + unsigned int len; + u32 state; + struct nsm_private priv; }; /* -- cgit v1.2.3-70-g09d2 From 3420a8c4359a189f7d854ed7075d151257415447 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:46 -0500 Subject: NSM: Add nsm_lookup() function Introduce a new API to fs/lockd/mon.c that allows nlm_host_rebooted() to lookup up nsm_handles via the contents of an nlm_reboot struct. The new function is equivalent to calling nsm_find() with @create set to zero, but it takes a struct nlm_reboot instead of separate arguments. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 64 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/lockd/lockd.h | 1 + 2 files changed, 65 insertions(+) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 4424b0a5a51..e46903995c9 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -201,6 +201,29 @@ void nsm_unmonitor(const struct nlm_host *host) } } +static struct nsm_handle *nsm_lookup_hostname(const char *hostname, + const size_t len) +{ + struct nsm_handle *nsm; + + list_for_each_entry(nsm, &nsm_handles, sm_link) + if (strlen(nsm->sm_name) == len && + memcmp(nsm->sm_name, hostname, len) == 0) + return nsm; + return NULL; +} + +static struct nsm_handle *nsm_lookup_priv(const struct nsm_private *priv) +{ + struct nsm_handle *nsm; + + list_for_each_entry(nsm, &nsm_handles, sm_link) + if (memcmp(nsm->sm_priv.data, priv->data, + sizeof(priv->data)) == 0) + return nsm; + return NULL; +} + /* * Construct a unique cookie to match this nsm_handle to this monitored * host. It is passed to the local rpc.statd via NSMPROC_MON, and @@ -297,6 +320,47 @@ found: return nsm; } +/** + * nsm_reboot_lookup - match NLMPROC_SM_NOTIFY arguments to an nsm_handle + * @info: pointer to NLMPROC_SM_NOTIFY arguments + * + * Returns a matching nsm_handle if found in the nsm cache; the returned + * nsm_handle's reference count is bumped and sm_monitored is cleared. + * Otherwise returns NULL if some error occurred. + */ +struct nsm_handle *nsm_reboot_lookup(const struct nlm_reboot *info) +{ + struct nsm_handle *cached; + + spin_lock(&nsm_lock); + + if (nsm_use_hostnames && info->mon != NULL) + cached = nsm_lookup_hostname(info->mon, info->len); + else + cached = nsm_lookup_priv(&info->priv); + + if (unlikely(cached == NULL)) { + spin_unlock(&nsm_lock); + dprintk("lockd: never saw rebooted peer '%.*s' before\n", + info->len, info->mon); + return cached; + } + + atomic_inc(&cached->sm_count); + spin_unlock(&nsm_lock); + + /* + * During subsequent lock activity, force a fresh + * notification to be set up for this host. + */ + cached->sm_monitored = 0; + + dprintk("lockd: host %s (%s) rebooted, cnt %d\n", + cached->sm_name, cached->sm_addrbuf, + atomic_read(&cached->sm_count)); + return cached; +} + /** * nsm_release - Release an NSM handle * @nsm: pointer to handle to be released diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 2a3533ea38d..5e3ad926de8 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -251,6 +251,7 @@ struct nsm_handle *nsm_find(const struct sockaddr *sap, const size_t salen, const char *hostname, const size_t hostname_len, const int create); +struct nsm_handle *nsm_reboot_lookup(const struct nlm_reboot *info); void nsm_release(struct nsm_handle *nsm); /* -- cgit v1.2.3-70-g09d2 From 8c7378fd2a5f22016542931b887a2ae98d146eaf Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:03:54 -0500 Subject: NLM: Call nsm_reboot_lookup() instead of nsm_find() Invoke the newly introduced nsm_reboot_lookup() function in nlm_host_rebooted() instead of nsm_find(). This introduces just one behavioral change: debugging messages produced during reboot notification will now appear when the NLMDBG_MONITOR flag is set, but not when the NLMDBG_HOSTCACHE flag is set. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 20 ++------------------ 1 file changed, 2 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index dc41e46ef74..230de93fc04 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -453,30 +453,14 @@ void nlm_release_host(struct nlm_host *host) */ void nlm_host_rebooted(const struct nlm_reboot *info) { - __be32 *p = (__be32 *)&info->priv.data; - const struct sockaddr_in sin = { - .sin_family = AF_INET, - .sin_addr.s_addr = *p, - }; struct hlist_head *chain; struct hlist_node *pos; struct nsm_handle *nsm; struct nlm_host *host; - nsm = nsm_find((struct sockaddr *)&sin, sizeof(sin), - info->mon, info->len, 0); - if (nsm == NULL) { - dprintk("lockd: never saw rebooted peer '%.*s' before\n", - info->len, info->mon); + nsm = nsm_reboot_lookup(info); + if (unlikely(nsm == NULL)) return; - } - - dprintk("lockd: nlm_host_rebooted(%.*s, %s)\n", - info->len, info->mon, nsm->sm_addrbuf); - - /* When reclaiming locks on this peer, make sure that - * we set up a new notification */ - nsm->sm_monitored = 0; /* Mark all hosts tied to this NSM state as having rebooted. * We run the loop repeatedly, because we drop the host table -- cgit v1.2.3-70-g09d2 From 92fd91b998a5216a6d6606704e71d541a180216c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 5 Dec 2008 19:04:01 -0500 Subject: NLM: Remove "create" argument from nsm_find() Clean up: nsm_find() now has only one caller, and that caller unconditionally sets the @create argument. Thus the @create argument is no longer needed. Since nsm_find() now has a more specific purpose, pick a more appropriate name for it. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/host.c | 4 ++-- fs/lockd/mon.c | 23 +++++++++-------------- include/linux/lockd/lockd.h | 6 +++--- 3 files changed, 14 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 230de93fc04..e5a65df4c0c 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -159,8 +159,8 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) atomic_inc(&nsm->sm_count); else { host = NULL; - nsm = nsm_find(ni->sap, ni->salen, - ni->hostname, ni->hostname_len, 1); + nsm = nsm_get_handle(ni->sap, ni->salen, + ni->hostname, ni->hostname_len); if (!nsm) { dprintk("lockd: nlm_lookup_host failed; " "no nsm handle\n"); diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index e46903995c9..74070221604 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -240,24 +240,22 @@ static void nsm_init_private(struct nsm_handle *nsm) } /** - * nsm_find - Find or create a cached nsm_handle + * nsm_get_handle - Find or create a cached nsm_handle * @sap: pointer to socket address of handle to find * @salen: length of socket address * @hostname: pointer to C string containing hostname to find * @hostname_len: length of C string - * @create: one means create new handle if not found in cache * - * Behavior is modulated by the global nsm_use_hostnames variable - * and by the @create argument. + * Behavior is modulated by the global nsm_use_hostnames variable. * - * Returns a cached nsm_handle after bumping its ref count, or if - * @create is set, returns a fresh nsm_handle if a handle that - * matches @sap and/or @hostname cannot be found in the handle cache. - * Returns NULL if an error occurs. + * Returns a cached nsm_handle after bumping its ref count, or + * returns a fresh nsm_handle if a handle that matches @sap and/or + * @hostname cannot be found in the handle cache. Returns NULL if + * an error occurs. */ -struct nsm_handle *nsm_find(const struct sockaddr *sap, const size_t salen, - const char *hostname, const size_t hostname_len, - const int create) +struct nsm_handle *nsm_get_handle(const struct sockaddr *sap, + const size_t salen, const char *hostname, + const size_t hostname_len) { struct nsm_handle *nsm = NULL; struct nsm_handle *pos; @@ -297,9 +295,6 @@ retry: } spin_unlock(&nsm_lock); - if (!create) - return NULL; - nsm = kzalloc(sizeof(*nsm) + hostname_len + 1, GFP_KERNEL); if (nsm == NULL) return NULL; diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 5e3ad926de8..1ccd49e97a7 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -247,10 +247,10 @@ void nlm_host_rebooted(const struct nlm_reboot *); int nsm_monitor(const struct nlm_host *host); void nsm_unmonitor(const struct nlm_host *host); -struct nsm_handle *nsm_find(const struct sockaddr *sap, const size_t salen, +struct nsm_handle *nsm_get_handle(const struct sockaddr *sap, + const size_t salen, const char *hostname, - const size_t hostname_len, - const int create); + const size_t hostname_len); struct nsm_handle *nsm_reboot_lookup(const struct nlm_reboot *info); void nsm_release(struct nsm_handle *nsm); -- cgit v1.2.3-70-g09d2 From b39b897c259fc1fd1998505f2b1d4ec1f115bce1 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:55:52 -0500 Subject: NSM: Refactor nsm_handle creation into a helper function Clean up. Refactor the creation of nsm_handles into a helper. Fields are initialized in increasing address order to make efficient use of CPU caches. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 38 ++++++++++++++++++++++++++------------ 1 file changed, 26 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 74070221604..315ca07715c 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -239,6 +239,30 @@ static void nsm_init_private(struct nsm_handle *nsm) *p = nsm_addr_in(nsm)->sin_addr.s_addr; } +static struct nsm_handle *nsm_create_handle(const struct sockaddr *sap, + const size_t salen, + const char *hostname, + const size_t hostname_len) +{ + struct nsm_handle *new; + + new = kzalloc(sizeof(*new) + hostname_len + 1, GFP_KERNEL); + if (unlikely(new == NULL)) + return NULL; + + atomic_set(&new->sm_count, 1); + new->sm_name = (char *)(new + 1); + memcpy(nsm_addr(new), sap, salen); + new->sm_addrlen = salen; + nsm_init_private(new); + nsm_display_address((const struct sockaddr *)&new->sm_addr, + new->sm_addrbuf, sizeof(new->sm_addrbuf)); + memcpy(new->sm_name, hostname, hostname_len); + new->sm_name[hostname_len] = '\0'; + + return new; +} + /** * nsm_get_handle - Find or create a cached nsm_handle * @sap: pointer to socket address of handle to find @@ -295,19 +319,9 @@ retry: } spin_unlock(&nsm_lock); - nsm = kzalloc(sizeof(*nsm) + hostname_len + 1, GFP_KERNEL); - if (nsm == NULL) + nsm = nsm_create_handle(sap, salen, hostname, hostname_len); + if (unlikely(nsm == NULL)) return NULL; - - memcpy(nsm_addr(nsm), sap, salen); - nsm->sm_addrlen = salen; - nsm->sm_name = (char *) (nsm + 1); - memcpy(nsm->sm_name, hostname, hostname_len); - nsm->sm_name[hostname_len] = '\0'; - nsm_init_private(nsm); - nsm_display_address((struct sockaddr *)&nsm->sm_addr, - nsm->sm_addrbuf, sizeof(nsm->sm_addrbuf)); - atomic_set(&nsm->sm_count, 1); goto retry; found: -- cgit v1.2.3-70-g09d2 From 77a3ef33e2de6fc8aabd7cb1700bfef81757c28a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:55:59 -0500 Subject: NSM: More clean up of nsm_get_handle() Clean up: refactor nsm_get_handle() so it is organized the same way that nsm_reboot_lookup() is. There is an additional micro-optimization here. This change moves the "hostname & nsm_use_hostnames" test out of the list_for_each_entry() clause in nsm_get_handle(), since it is loop-invariant. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 62 +++++++++++++++++++++++++++++++++------------------------- 1 file changed, 35 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 315ca07715c..99aec744474 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -213,6 +213,16 @@ static struct nsm_handle *nsm_lookup_hostname(const char *hostname, return NULL; } +static struct nsm_handle *nsm_lookup_addr(const struct sockaddr *sap) +{ + struct nsm_handle *nsm; + + list_for_each_entry(nsm, &nsm_handles, sm_link) + if (nlm_cmp_addr(nsm_addr(nsm), sap)) + return nsm; + return NULL; +} + static struct nsm_handle *nsm_lookup_priv(const struct nsm_private *priv) { struct nsm_handle *nsm; @@ -281,8 +291,7 @@ struct nsm_handle *nsm_get_handle(const struct sockaddr *sap, const size_t salen, const char *hostname, const size_t hostname_len) { - struct nsm_handle *nsm = NULL; - struct nsm_handle *pos; + struct nsm_handle *cached, *new = NULL; if (hostname && memchr(hostname, '/', hostname_len) != NULL) { if (printk_ratelimit()) { @@ -295,38 +304,37 @@ struct nsm_handle *nsm_get_handle(const struct sockaddr *sap, retry: spin_lock(&nsm_lock); - list_for_each_entry(pos, &nsm_handles, sm_link) { - - if (hostname && nsm_use_hostnames) { - if (strlen(pos->sm_name) != hostname_len - || memcmp(pos->sm_name, hostname, hostname_len)) - continue; - } else if (!nlm_cmp_addr(nsm_addr(pos), sap)) - continue; - atomic_inc(&pos->sm_count); - kfree(nsm); - nsm = pos; - dprintk("lockd: found nsm_handle for %s (%s), cnt %d\n", - pos->sm_name, pos->sm_addrbuf, - atomic_read(&pos->sm_count)); - goto found; + + if (nsm_use_hostnames && hostname != NULL) + cached = nsm_lookup_hostname(hostname, hostname_len); + else + cached = nsm_lookup_addr(sap); + + if (cached != NULL) { + atomic_inc(&cached->sm_count); + spin_unlock(&nsm_lock); + kfree(new); + dprintk("lockd: found nsm_handle for %s (%s), " + "cnt %d\n", cached->sm_name, + cached->sm_addrbuf, + atomic_read(&cached->sm_count)); + return cached; } - if (nsm) { - list_add(&nsm->sm_link, &nsm_handles); + + if (new != NULL) { + list_add(&new->sm_link, &nsm_handles); + spin_unlock(&nsm_lock); dprintk("lockd: created nsm_handle for %s (%s)\n", - nsm->sm_name, nsm->sm_addrbuf); - goto found; + new->sm_name, new->sm_addrbuf); + return new; } + spin_unlock(&nsm_lock); - nsm = nsm_create_handle(sap, salen, hostname, hostname_len); - if (unlikely(nsm == NULL)) + new = nsm_create_handle(sap, salen, hostname, hostname_len); + if (unlikely(new == NULL)) return NULL; goto retry; - -found: - spin_unlock(&nsm_lock); - return nsm; } /** -- cgit v1.2.3-70-g09d2 From 94da7663db26530a8377f7219f8be8bd4d4822c2 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:56:07 -0500 Subject: NSM: Replace IP address as our nlm_reboot lookup key NLM provides file locking services for NFS files. Part of this service includes a second protocol, known as NSM, which is a reboot notification service. NLM uses this service to determine when to reclaim locks or enter a grace period after a client or server reboots. The NLM service (implemented by lockd in the Linux kernel) contacts the local NSM service (implemented by rpc.statd in Linux user space) via NSM protocol upcalls to register a callback when a particular remote peer reboots. To match the callback to the correct remote peer, the NLM service constructs a cookie that it passes in the request. The NSM service passes that cookie back to the NLM service when it is notified that the given remote peer has indeed rebooted. Currently on Linux, the cookie is the raw 32-bit IPv4 address of the remote peer. To support IPv6 addresses, which are larger, we could use all 16 bytes of the cookie to represent a full IPv6 address, although we still can't represent an IPv6 address with a scope ID in just 16 bytes. Instead, to avoid the need for future changes to support additional address types, we'll use a manufactured value for the cookie, and use that to find the corresponding nsm_handle struct in the kernel during the NLMPROC_SM_NOTIFY callback. This should provide complete support in the kernel's NSM implementation for IPv6 hosts, while remaining backwards compatible with older rpc.statd implementations. Note we also deal with another case where nsm_use_hostnames can change while there are outstanding notifications, possibly resulting in the loss of reboot notifications. After this patch, the priv cookie is always used to lookup rebooted hosts in the kernel. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 28 +++++++++++++++++++--------- 1 file changed, 19 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 99aec744474..8ae4c02d7df 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -9,6 +9,8 @@ #include #include #include +#include + #include #include #include @@ -240,13 +242,25 @@ static struct nsm_handle *nsm_lookup_priv(const struct nsm_private *priv) * returned via NLMPROC_SM_NOTIFY, in the "priv" field of these * requests. * - * Linux provides the raw IP address of the monitored host, - * left in network byte order. + * The NSM protocol requires that these cookies be unique while the + * system is running. We prefer a stronger requirement of making them + * unique across reboots. If user space bugs cause a stale cookie to + * be sent to the kernel, it could cause the wrong host to lose its + * lock state if cookies were not unique across reboots. + * + * The cookies are exposed only to local user space via loopback. They + * do not appear on the physical network. If we want greater security + * for some reason, nsm_init_private() could perform a one-way hash to + * obscure the contents of the cookie. */ static void nsm_init_private(struct nsm_handle *nsm) { - __be32 *p = (__be32 *)&nsm->sm_priv.data; - *p = nsm_addr_in(nsm)->sin_addr.s_addr; + u64 *p = (u64 *)&nsm->sm_priv.data; + struct timespec ts; + + ktime_get_ts(&ts); + *p++ = timespec_to_ns(&ts); + *p = (unsigned long)nsm; } static struct nsm_handle *nsm_create_handle(const struct sockaddr *sap, @@ -351,11 +365,7 @@ struct nsm_handle *nsm_reboot_lookup(const struct nlm_reboot *info) spin_lock(&nsm_lock); - if (nsm_use_hostnames && info->mon != NULL) - cached = nsm_lookup_hostname(info->mon, info->len); - else - cached = nsm_lookup_priv(&info->priv); - + cached = nsm_lookup_priv(&info->priv); if (unlikely(cached == NULL)) { spin_unlock(&nsm_lock); dprintk("lockd: never saw rebooted peer '%.*s' before\n", -- cgit v1.2.3-70-g09d2 From e6765b83977f07983c7a10e6bbb19d6c7bbfc3a4 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:56:14 -0500 Subject: NSM: Remove include/linux/lockd/sm_inter.h Clean up: The include/linux/lockd/sm_inter.h header is nearly empty now. Remove it. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/clntproc.c | 1 - fs/lockd/host.c | 1 - fs/lockd/mon.c | 2 -- fs/lockd/svc.c | 1 - fs/lockd/svc4proc.c | 2 -- fs/lockd/svcproc.c | 2 -- fs/lockd/svcsubs.c | 1 - fs/lockd/xdr.c | 1 - fs/lockd/xdr4.c | 1 - include/linux/lockd/lockd.h | 1 + include/linux/lockd/sm_inter.h | 16 ---------------- include/linux/lockd/xdr.h | 1 + 12 files changed, 2 insertions(+), 28 deletions(-) delete mode 100644 include/linux/lockd/sm_inter.h (limited to 'fs') diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 5ce42e0ed4a..dd7957064a8 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -16,7 +16,6 @@ #include #include #include -#include #define NLMDBG_FACILITY NLMDBG_CLIENT #define NLMCLNT_GRACE_WAIT (5*HZ) diff --git a/fs/lockd/host.c b/fs/lockd/host.c index e5a65df4c0c..99d737bd432 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 8ae4c02d7df..dfa9d80efcb 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -15,8 +15,6 @@ #include #include #include -#include - #define NLMDBG_FACILITY NLMDBG_MONITOR #define NSM_PROGRAM 100024 diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index bc3c3cb62db..0b13392931a 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -35,7 +35,6 @@ #include #include #include -#include #include #define NLMDBG_FACILITY NLMDBG_SVC diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c index bb79a53e060..1725037374c 100644 --- a/fs/lockd/svc4proc.c +++ b/fs/lockd/svc4proc.c @@ -16,8 +16,6 @@ #include #include #include -#include - #define NLMDBG_FACILITY NLMDBG_CLIENT diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index e44310c0211..3688e55901f 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -16,8 +16,6 @@ #include #include #include -#include - #define NLMDBG_FACILITY NLMDBG_CLIENT diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index 34c2766e27c..9e4d6aab611 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include diff --git a/fs/lockd/xdr.c b/fs/lockd/xdr.c index 4cc7d01a1eb..0336f2beacd 100644 --- a/fs/lockd/xdr.c +++ b/fs/lockd/xdr.c @@ -16,7 +16,6 @@ #include #include #include -#include #define NLMDBG_FACILITY NLMDBG_XDR diff --git a/fs/lockd/xdr4.c b/fs/lockd/xdr4.c index 61d1714a470..e1d52865319 100644 --- a/fs/lockd/xdr4.c +++ b/fs/lockd/xdr4.c @@ -17,7 +17,6 @@ #include #include #include -#include #define NLMDBG_FACILITY NLMDBG_XDR diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 1ccd49e97a7..8b57467375c 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -205,6 +205,7 @@ extern struct svc_procedure nlmsvc_procedures4[]; extern int nlmsvc_grace_period; extern unsigned long nlmsvc_timeout; extern int nsm_use_hostnames; +extern int nsm_local_state; /* * Lockd client functions diff --git a/include/linux/lockd/sm_inter.h b/include/linux/lockd/sm_inter.h deleted file mode 100644 index 5cef5a79dd9..00000000000 --- a/include/linux/lockd/sm_inter.h +++ /dev/null @@ -1,16 +0,0 @@ -/* - * linux/include/linux/lockd/sm_inter.h - * - * Declarations for the kernel statd client. - * - * Copyright (C) 1996, Olaf Kirch - */ - -#ifndef LINUX_LOCKD_SM_INTER_H -#define LINUX_LOCKD_SM_INTER_H - -#define SM_MAXSTRLEN 1024 - -extern int nsm_local_state; - -#endif /* LINUX_LOCKD_SM_INTER_H */ diff --git a/include/linux/lockd/xdr.h b/include/linux/lockd/xdr.h index 6338866222a..7dc5b6cb44c 100644 --- a/include/linux/lockd/xdr.h +++ b/include/linux/lockd/xdr.h @@ -13,6 +13,7 @@ #include #include +#define SM_MAXSTRLEN 1024 #define SM_PRIV_SIZE 16 struct nsm_private { -- cgit v1.2.3-70-g09d2 From 8529bc51d30b8f001734b29b21a51b579c260f5b Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:56:22 -0500 Subject: NSM: Move nsm_addr() to fs/lockd/mon.c Clean up: nsm_addr_in() is no longer used, and nsm_addr() is used only in fs/lockd/mon.c, so move it there. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 5 +++++ include/linux/lockd/lockd.h | 10 ---------- 2 files changed, 5 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index dfa9d80efcb..43be31c4a2d 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -55,6 +55,11 @@ static DEFINE_SPINLOCK(nsm_lock); */ int nsm_local_state; +static inline struct sockaddr *nsm_addr(const struct nsm_handle *nsm) +{ + return (struct sockaddr *)&nsm->sm_addr; +} + static void nsm_display_ipv4_address(const struct sockaddr *sap, char *buf, const size_t len) { diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 8b57467375c..6ab0449bc82 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -112,16 +112,6 @@ static inline struct sockaddr *nlm_srcaddr(const struct nlm_host *host) return (struct sockaddr *)&host->h_srcaddr; } -static inline struct sockaddr_in *nsm_addr_in(const struct nsm_handle *handle) -{ - return (struct sockaddr_in *)&handle->sm_addr; -} - -static inline struct sockaddr *nsm_addr(const struct nsm_handle *handle) -{ - return (struct sockaddr *)&handle->sm_addr; -} - /* * Map an fl_owner_t into a unique 32-bit "pid" */ -- cgit v1.2.3-70-g09d2 From b7ba597fb964dfa44284904b3b3d74d44b8e1c42 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:56:29 -0500 Subject: NSM: Move nsm_use_hostnames to mon.c Clean up. Treat the nsm_use_hostnames global variable like nsm_local_state. Note that the default value of nsm_use_hostnames is still zero. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 3 ++- fs/lockd/svc.c | 1 - 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 43be31c4a2d..fafa0ea7193 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -53,7 +53,8 @@ static DEFINE_SPINLOCK(nsm_lock); /* * Local NSM state */ -int nsm_local_state; +int __read_mostly nsm_local_state; +int __read_mostly nsm_use_hostnames; static inline struct sockaddr *nsm_addr(const struct nsm_handle *nsm) { diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 0b13392931a..7ac7d72e3b5 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -59,7 +59,6 @@ unsigned long nlmsvc_timeout; static unsigned long nlm_grace_period; static unsigned long nlm_timeout = LOCKD_DFLT_TIMEO; static int nlm_udpport, nlm_tcpport; -int nsm_use_hostnames = 0; /* RLIM_NOFILE defaults to 1024. That seems like a reasonable default here. */ static unsigned int nlm_max_connections = 1024; -- cgit v1.2.3-70-g09d2 From 49b5699b3fc22b363534c509c1b7dba06bc677bf Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:56:37 -0500 Subject: NSM: Move nsm_create() Clean up: one last thing... relocate nsm_create() to eliminate the forward declaration and group it near the only function that actually uses it. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/mon.c | 51 ++++++++++++++++++++------------------------------- 1 file changed, 20 insertions(+), 31 deletions(-) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index fafa0ea7193..5e2c4d5ac82 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -44,8 +44,6 @@ struct nsm_res { u32 state; }; -static struct rpc_clnt * nsm_create(void); - static struct rpc_program nsm_program; static LIST_HEAD(nsm_handles); static DEFINE_SPINLOCK(nsm_lock); @@ -98,11 +96,26 @@ static void nsm_display_address(const struct sockaddr *sap, } } -/* - * Common procedure for NSMPROC_MON/NSMPROC_UNMON calls - */ -static int -nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) +static struct rpc_clnt *nsm_create(void) +{ + struct sockaddr_in sin = { + .sin_family = AF_INET, + .sin_addr.s_addr = htonl(INADDR_LOOPBACK), + }; + struct rpc_create_args args = { + .protocol = XPRT_TRANSPORT_UDP, + .address = (struct sockaddr *)&sin, + .addrsize = sizeof(sin), + .servername = "rpc.statd", + .program = &nsm_program, + .version = NSM_VERSION, + .authflavor = RPC_AUTH_NULL, + }; + + return rpc_create(&args); +} + +static int nsm_mon_unmon(struct nsm_handle *nsm, u32 proc, struct nsm_res *res) { struct rpc_clnt *clnt; int status; @@ -408,30 +421,6 @@ void nsm_release(struct nsm_handle *nsm) } } -/* - * Create NSM client for the local host - */ -static struct rpc_clnt * -nsm_create(void) -{ - struct sockaddr_in sin = { - .sin_family = AF_INET, - .sin_addr.s_addr = htonl(INADDR_LOOPBACK), - .sin_port = 0, - }; - struct rpc_create_args args = { - .protocol = XPRT_TRANSPORT_UDP, - .address = (struct sockaddr *)&sin, - .addrsize = sizeof(sin), - .servername = "localhost", - .program = &nsm_program, - .version = NSM_VERSION, - .authflavor = RPC_AUTH_NULL, - }; - - return rpc_create(&args); -} - /* * XDR functions for NSM. * -- cgit v1.2.3-70-g09d2 From b064ec038a6180b13e5f89b6a30b42cb5ce8febc Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Dec 2008 17:56:59 -0500 Subject: lockd: Enable NLM use of AF_INET6 If the kernel is configured to support IPv6 and the RPC server can register services via rpcbindv4, we are all set to enable IPv6 support for lockd. Signed-off-by: Chuck Lever Cc: Aime Le Rouzic Signed-off-by: J. Bruce Fields --- fs/lockd/svc.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 7ac7d72e3b5..3e5f9f07911 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -52,6 +52,17 @@ static struct task_struct *nlmsvc_task; static struct svc_rqst *nlmsvc_rqst; unsigned long nlmsvc_timeout; +/* + * If the kernel has IPv6 support available, always listen for + * both AF_INET and AF_INET6 requests. + */ +#if (defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)) && \ + defined(CONFIG_SUNRPC_REGISTER_V4) +static const sa_family_t nlmsvc_family = AF_INET6; +#else /* (CONFIG_IPV6 || CONFIG_IPV6_MODULE) && CONFIG_SUNRPC_REGISTER_V4 */ +static const sa_family_t nlmsvc_family = AF_INET; +#endif /* (CONFIG_IPV6 || CONFIG_IPV6_MODULE) && CONFIG_SUNRPC_REGISTER_V4 */ + /* * These can be set at insmod time (useful for NFS as root filesystem), * and also changed through the sysctl interface. -- Jamie Lokier, Aug 2003 @@ -256,7 +267,7 @@ int lockd_up(void) "lockd_up: no pid, %d users??\n", nlmsvc_users); error = -ENOMEM; - serv = svc_create(&nlmsvc_program, LOCKD_BUFSIZE, AF_INET, NULL); + serv = svc_create(&nlmsvc_program, LOCKD_BUFSIZE, nlmsvc_family, NULL); if (!serv) { printk(KERN_WARNING "lockd_up: create service failed\n"); goto out; -- cgit v1.2.3-70-g09d2 From b046ccdc1f8171f6d0129dcc2a28d49187b4bf69 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 12 Dec 2008 16:57:13 -0500 Subject: NFSD: clean up failover sysctl function naming Clean up: Rename recently-added failover functions to match the naming convention in fs/nfsd/nfsctl.c. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 77d7b8c531a..173c4dd3d7a 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -84,6 +84,8 @@ static ssize_t write_unexport(struct file *file, char *buf, size_t size); static ssize_t write_getfd(struct file *file, char *buf, size_t size); static ssize_t write_getfs(struct file *file, char *buf, size_t size); static ssize_t write_filehandle(struct file *file, char *buf, size_t size); +static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size); +static ssize_t write_unlock_fs(struct file *file, char *buf, size_t size); static ssize_t write_threads(struct file *file, char *buf, size_t size); static ssize_t write_pool_threads(struct file *file, char *buf, size_t size); static ssize_t write_versions(struct file *file, char *buf, size_t size); @@ -94,9 +96,6 @@ static ssize_t write_leasetime(struct file *file, char *buf, size_t size); static ssize_t write_recoverydir(struct file *file, char *buf, size_t size); #endif -static ssize_t failover_unlock_ip(struct file *file, char *buf, size_t size); -static ssize_t failover_unlock_fs(struct file *file, char *buf, size_t size); - static ssize_t (*write_op[])(struct file *, char *, size_t) = { [NFSD_Svc] = write_svc, [NFSD_Add] = write_add, @@ -106,8 +105,8 @@ static ssize_t (*write_op[])(struct file *, char *, size_t) = { [NFSD_Getfd] = write_getfd, [NFSD_Getfs] = write_getfs, [NFSD_Fh] = write_filehandle, - [NFSD_FO_UnlockIP] = failover_unlock_ip, - [NFSD_FO_UnlockFS] = failover_unlock_fs, + [NFSD_FO_UnlockIP] = write_unlock_ip, + [NFSD_FO_UnlockFS] = write_unlock_fs, [NFSD_Threads] = write_threads, [NFSD_Pool_Threads] = write_pool_threads, [NFSD_Versions] = write_versions, @@ -309,7 +308,7 @@ static ssize_t write_getfd(struct file *file, char *buf, size_t size) return err; } -static ssize_t failover_unlock_ip(struct file *file, char *buf, size_t size) +static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size) { struct sockaddr_in sin = { .sin_family = AF_INET, @@ -339,7 +338,7 @@ static ssize_t failover_unlock_ip(struct file *file, char *buf, size_t size) return nlmsvc_unlock_all_by_ip((struct sockaddr *)&sin); } -static ssize_t failover_unlock_fs(struct file *file, char *buf, size_t size) +static ssize_t write_unlock_fs(struct file *file, char *buf, size_t size) { struct path path; char *fo_path; -- cgit v1.2.3-70-g09d2 From 54224f04ae95d86b27c0673cd773ebb120d86876 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 12 Dec 2008 16:57:20 -0500 Subject: NFSD: Fix a handful of coding style issues in write_filehandle() Clean up: follow kernel coding style. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 173c4dd3d7a..498d763b932 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -390,11 +390,13 @@ static ssize_t write_filehandle(struct file *file, char *buf, size_t size) dname = mesg; len = qword_get(&mesg, dname, size); - if (len <= 0) return -EINVAL; + if (len <= 0) + return -EINVAL; path = dname+len+1; len = qword_get(&mesg, path, size); - if (len <= 0) return -EINVAL; + if (len <= 0) + return -EINVAL; len = get_int(&mesg, &maxsize); if (len) @@ -418,7 +420,8 @@ static ssize_t write_filehandle(struct file *file, char *buf, size_t size) if (len) return len; - mesg = buf; len = SIMPLE_TRANSACTION_LIMIT; + mesg = buf; + len = SIMPLE_TRANSACTION_LIMIT; qword_addhex(&mesg, &len, (char*)&fh.fh_base, fh.fh_size); mesg[-1] = '\n'; return mesg - buf; -- cgit v1.2.3-70-g09d2 From 9e074856caf13ba83363f73759f5e395f74ccf41 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 12 Dec 2008 16:57:27 -0500 Subject: NFSD: Replace open-coded integer with macro Clean up: Instead of open-coding 2049, use the NFS_PORT macro. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 498d763b932..856b8646a48 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -439,9 +439,9 @@ static ssize_t write_threads(struct file *file, char *buf, size_t size) rv = get_int(&mesg, &newthreads); if (rv) return rv; - if (newthreads <0) + if (newthreads < 0) return -EINVAL; - rv = nfsd_svc(2049, newthreads); + rv = nfsd_svc(NFS_PORT, newthreads); if (rv) return rv; } -- cgit v1.2.3-70-g09d2 From 262a09823bb07c6aafb6c1d312cde613d0b90c85 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 12 Dec 2008 16:57:35 -0500 Subject: NFSD: Add documenting comments for nfsctl interface Document the NFSD sysctl interface laid out in fs/nfsd/nfsctl.c. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 453 +++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 437 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 856b8646a48..3d93b2064ce 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -175,10 +175,24 @@ static const struct file_operations exports_operations = { /*----------------------------------------------------------------------------*/ /* * payload - write methods - * If the method has a response, the response should be put in buf, - * and the length returned. Otherwise return 0 or and -error. */ +/** + * write_svc - Start kernel's NFSD server + * + * Deprecated. /proc/fs/nfsd/threads is preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_svc + * svc_port: port number of this + * server's listener + * svc_nthreads: number of threads to start + * size: size in bytes of passed in nfsctl_svc + * Output: + * On success: returns zero + * On error: return code is negative errno value + */ static ssize_t write_svc(struct file *file, char *buf, size_t size) { struct nfsctl_svc *data; @@ -188,6 +202,30 @@ static ssize_t write_svc(struct file *file, char *buf, size_t size) return nfsd_svc(data->svc_port, data->svc_nthreads); } +/** + * write_add - Add or modify client entry in auth unix cache + * + * Deprecated. /proc/net/rpc/auth.unix.ip is preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_client + * cl_ident: '\0'-terminated C string + * containing domain name + * of client + * cl_naddr: no. of items in cl_addrlist + * cl_addrlist: array of client addresses + * cl_fhkeytype: ignored + * cl_fhkeylen: ignored + * cl_fhkey: ignored + * size: size in bytes of passed in nfsctl_client + * Output: + * On success: returns zero + * On error: return code is negative errno value + * + * Note: Only AF_INET client addresses are passed in, since + * nfsctl_client.cl_addrlist contains only in_addr fields for addresses. + */ static ssize_t write_add(struct file *file, char *buf, size_t size) { struct nfsctl_client *data; @@ -197,6 +235,30 @@ static ssize_t write_add(struct file *file, char *buf, size_t size) return exp_addclient(data); } +/** + * write_del - Remove client from auth unix cache + * + * Deprecated. /proc/net/rpc/auth.unix.ip is preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_client + * cl_ident: '\0'-terminated C string + * containing domain name + * of client + * cl_naddr: ignored + * cl_addrlist: ignored + * cl_fhkeytype: ignored + * cl_fhkeylen: ignored + * cl_fhkey: ignored + * size: size in bytes of passed in nfsctl_client + * Output: + * On success: returns zero + * On error: return code is negative errno value + * + * Note: Only AF_INET client addresses are passed in, since + * nfsctl_client.cl_addrlist contains only in_addr fields for addresses. + */ static ssize_t write_del(struct file *file, char *buf, size_t size) { struct nfsctl_client *data; @@ -206,6 +268,33 @@ static ssize_t write_del(struct file *file, char *buf, size_t size) return exp_delclient(data); } +/** + * write_export - Export part or all of a local file system + * + * Deprecated. /proc/net/rpc/{nfsd.export,nfsd.fh} are preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_export + * ex_client: '\0'-terminated C string + * containing domain name + * of client allowed to access + * this export + * ex_path: '\0'-terminated C string + * containing pathname of + * directory in local file system + * ex_dev: fsid to use for this export + * ex_ino: ignored + * ex_flags: export flags for this export + * ex_anon_uid: UID to use for anonymous + * requests + * ex_anon_gid: GID to use for anonymous + * requests + * size: size in bytes of passed in nfsctl_export + * Output: + * On success: returns zero + * On error: return code is negative errno value + */ static ssize_t write_export(struct file *file, char *buf, size_t size) { struct nfsctl_export *data; @@ -215,6 +304,31 @@ static ssize_t write_export(struct file *file, char *buf, size_t size) return exp_export(data); } +/** + * write_unexport - Unexport a previously exported file system + * + * Deprecated. /proc/net/rpc/{nfsd.export,nfsd.fh} are preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_export + * ex_client: '\0'-terminated C string + * containing domain name + * of client no longer allowed + * to access this export + * ex_path: '\0'-terminated C string + * containing pathname of + * directory in local file system + * ex_dev: ignored + * ex_ino: ignored + * ex_flags: ignored + * ex_anon_uid: ignored + * ex_anon_gid: ignored + * size: size in bytes of passed in nfsctl_export + * Output: + * On success: returns zero + * On error: return code is negative errno value + */ static ssize_t write_unexport(struct file *file, char *buf, size_t size) { struct nfsctl_export *data; @@ -225,6 +339,30 @@ static ssize_t write_unexport(struct file *file, char *buf, size_t size) return exp_unexport(data); } +/** + * write_getfs - Get a variable-length NFS file handle by path + * + * Deprecated. /proc/fs/nfsd/filehandle is preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_fsparm + * gd_addr: socket address of client + * gd_path: '\0'-terminated C string + * containing pathname of + * directory in local file system + * gd_maxlen: maximum size of returned file + * handle + * size: size in bytes of passed in nfsctl_fsparm + * Output: + * On success: passed-in buffer filled with a knfsd_fh structure + * (a variable-length raw NFS file handle); + * return code is the size in bytes of the file handle + * On error: return code is negative errno value + * + * Note: Only AF_INET client addresses are passed in, since gd_addr + * is the same size as a struct sockaddr_in. + */ static ssize_t write_getfs(struct file *file, char *buf, size_t size) { struct nfsctl_fsparm *data; @@ -264,6 +402,29 @@ static ssize_t write_getfs(struct file *file, char *buf, size_t size) return err; } +/** + * write_getfd - Get a fixed-length NFS file handle by path (used by mountd) + * + * Deprecated. /proc/fs/nfsd/filehandle is preferred. + * Function remains to support old versions of nfs-utils. + * + * Input: + * buf: struct nfsctl_fdparm + * gd_addr: socket address of client + * gd_path: '\0'-terminated C string + * containing pathname of + * directory in local file system + * gd_version: fdparm structure version + * size: size in bytes of passed in nfsctl_fdparm + * Output: + * On success: passed-in buffer filled with nfsctl_res + * (a fixed-length raw NFS file handle); + * return code is the size in bytes of the file handle + * On error: return code is negative errno value + * + * Note: Only AF_INET client addresses are passed in, since gd_addr + * is the same size as a struct sockaddr_in. + */ static ssize_t write_getfd(struct file *file, char *buf, size_t size) { struct nfsctl_fdparm *data; @@ -308,6 +469,22 @@ static ssize_t write_getfd(struct file *file, char *buf, size_t size) return err; } +/** + * write_unlock_ip - Release all locks used by a client + * + * Experimental. + * + * Input: + * buf: '\n'-terminated C string containing a + * presentation format IPv4 address + * size: length of C string in @buf + * Output: + * On success: returns zero if all specified locks were released; + * returns one if one or more locks were not released + * On error: return code is negative errno value + * + * Note: Only AF_INET client addresses are passed in + */ static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size) { struct sockaddr_in sin = { @@ -338,6 +515,20 @@ static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size) return nlmsvc_unlock_all_by_ip((struct sockaddr *)&sin); } +/** + * write_unlock_fs - Release all locks on a local file system + * + * Experimental. + * + * Input: + * buf: '\n'-terminated C string containing the + * absolute pathname of a local file system + * size: length of C string in @buf + * Output: + * On success: returns zero if all specified locks were released; + * returns one if one or more locks were not released + * On error: return code is negative errno value + */ static ssize_t write_unlock_fs(struct file *file, char *buf, size_t size) { struct path path; @@ -359,21 +550,44 @@ static ssize_t write_unlock_fs(struct file *file, char *buf, size_t size) if (error) return error; + /* + * XXX: Needs better sanity checking. Otherwise we could end up + * releasing locks on the wrong file system. + * + * For example: + * 1. Does the path refer to a directory? + * 2. Is that directory a mount point, or + * 3. Is that directory the root of an exported file system? + */ error = nlmsvc_unlock_all_by_sb(path.mnt->mnt_sb); path_put(&path); return error; } +/** + * write_filehandle - Get a variable-length NFS file handle by path + * + * On input, the buffer contains a '\n'-terminated C string comprised of + * three alphanumeric words separated by whitespace. The string may + * contain escape sequences. + * + * Input: + * buf: + * domain: client domain name + * path: export pathname + * maxsize: numeric maximum size of + * @buf + * size: length of C string in @buf + * Output: + * On success: passed-in buffer filled with '\n'-terminated C + * string containing a ASCII hex text version + * of the NFS file handle; + * return code is the size in bytes of the string + * On error: return code is negative errno value + */ static ssize_t write_filehandle(struct file *file, char *buf, size_t size) { - /* request is: - * domain path maxsize - * response is - * filehandle - * - * qword quoting is used, so filehandle will be \x.... - */ char *dname, *path; int uninitialized_var(maxsize); char *mesg = buf; @@ -427,11 +641,36 @@ static ssize_t write_filehandle(struct file *file, char *buf, size_t size) return mesg - buf; } +/** + * write_threads - Start NFSD, or report the current number of running threads + * + * Input: + * buf: ignored + * size: zero + * Output: + * On success: passed-in buffer filled with '\n'-terminated C + * string numeric value representing the number of + * running NFSD threads; + * return code is the size in bytes of the string + * On error: return code is zero + * + * OR + * + * Input: + * buf: C string containing an unsigned + * integer value representing the + * number of NFSD threads to start + * size: non-zero length of C string in @buf + * Output: + * On success: NFS service is started; + * passed-in buffer filled with '\n'-terminated C + * string numeric value representing the number of + * running NFSD threads; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + */ static ssize_t write_threads(struct file *file, char *buf, size_t size) { - /* if size > 0, look for a number of threads and call nfsd_svc - * then write out number of threads as reply - */ char *mesg = buf; int rv; if (size > 0) { @@ -449,6 +688,28 @@ static ssize_t write_threads(struct file *file, char *buf, size_t size) return strlen(buf); } +/** + * write_pool_threads - Set or report the current number of threads per pool + * + * Input: + * buf: ignored + * size: zero + * + * OR + * + * Input: + * buf: C string containing whitespace- + * separated unsigned integer values + * representing the number of NFSD + * threads to start in each pool + * size: non-zero length of C string in @buf + * Output: + * On success: passed-in buffer filled with '\n'-terminated C + * string containing integer values representing the + * number of NFSD threads in each pool; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + */ static ssize_t write_pool_threads(struct file *file, char *buf, size_t size) { /* if size > 0, look for an array of number of threads per node @@ -519,10 +780,6 @@ out_free: static ssize_t __write_versions(struct file *file, char *buf, size_t size) { - /* - * Format: - * [-/+]vers [-/+]vers ... - */ char *mesg = buf; char *vers, sign; int len, num; @@ -580,6 +837,38 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) return len; } +/** + * write_versions - Set or report the available NFS protocol versions + * + * Input: + * buf: ignored + * size: zero + * Output: + * On success: passed-in buffer filled with '\n'-terminated C + * string containing positive or negative integer + * values representing the current status of each + * protocol version; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + * + * OR + * + * Input: + * buf: C string containing whitespace- + * separated positive or negative + * integer values representing NFS + * protocol versions to enable ("+n") + * or disable ("-n") + * size: non-zero length of C string in @buf + * Output: + * On success: status of zero or more protocol versions has + * been updated; passed-in buffer filled with + * '\n'-terminated C string containing positive + * or negative integer values representing the + * current status of each protocol version; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + */ static ssize_t write_versions(struct file *file, char *buf, size_t size) { ssize_t rv; @@ -689,6 +978,75 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size) return -EINVAL; } +/** + * write_ports - Pass a socket file descriptor or transport name to listen on + * + * Input: + * buf: ignored + * size: zero + * Output: + * On success: passed-in buffer filled with a '\n'-terminated C + * string containing a whitespace-separated list of + * named NFSD listeners; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + * + * OR + * + * Input: + * buf: C string containing an unsigned + * integer value representing a bound + * but unconnected socket that is to be + * used as an NFSD listener + * size: non-zero length of C string in @buf + * Output: + * On success: NFS service is started; + * passed-in buffer filled with a '\n'-terminated C + * string containing a unique alphanumeric name of + * the listener; + * return code is the size in bytes of the string + * On error: return code is a negative errno value + * + * OR + * + * Input: + * buf: C string containing a "-" followed + * by an integer value representing a + * previously passed in socket file + * descriptor + * size: non-zero length of C string in @buf + * Output: + * On success: NFS service no longer listens on that socket; + * passed-in buffer filled with a '\n'-terminated C + * string containing a unique name of the listener; + * return code is the size in bytes of the string + * On error: return code is a negative errno value + * + * OR + * + * Input: + * buf: C string containing a transport + * name and an unsigned integer value + * representing the port to listen on, + * separated by whitespace + * size: non-zero length of C string in @buf + * Output: + * On success: returns zero; NFS service is started + * On error: return code is a negative errno value + * + * OR + * + * Input: + * buf: C string containing a "-" followed + * by a transport name and an unsigned + * integer value representing the port + * to listen on, separated by whitespace + * size: non-zero length of C string in @buf + * Output: + * On success: returns zero; NFS service no longer listens + * on that transport + * On error: return code is a negative errno value + */ static ssize_t write_ports(struct file *file, char *buf, size_t size) { ssize_t rv; @@ -702,6 +1060,27 @@ static ssize_t write_ports(struct file *file, char *buf, size_t size) int nfsd_max_blksize; +/** + * write_maxblksize - Set or report the current NFS blksize + * + * Input: + * buf: ignored + * size: zero + * + * OR + * + * Input: + * buf: C string containing an unsigned + * integer value representing the new + * NFS blksize + * size: non-zero length of C string in @buf + * Output: + * On success: passed-in buffer filled with '\n'-terminated C string + * containing numeric value of the current NFS blksize + * setting; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + */ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) { char *mesg = buf; @@ -754,6 +1133,27 @@ static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) return strlen(buf); } +/** + * write_leasetime - Set or report the current NFSv4 lease time + * + * Input: + * buf: ignored + * size: zero + * + * OR + * + * Input: + * buf: C string containing an unsigned + * integer value representing the new + * NFSv4 lease expiry time + * size: non-zero length of C string in @buf + * Output: + * On success: passed-in buffer filled with '\n'-terminated C + * string containing unsigned integer value of the + * current lease expiry time; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + */ static ssize_t write_leasetime(struct file *file, char *buf, size_t size) { ssize_t rv; @@ -790,6 +1190,27 @@ static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size) return strlen(buf); } +/** + * write_recoverydir - Set or report the pathname of the recovery directory + * + * Input: + * buf: ignored + * size: zero + * + * OR + * + * Input: + * buf: C string containing the pathname + * of the directory on a local file + * system containing permanent NFSv4 + * recovery data + * size: non-zero length of C string in @buf + * Output: + * On success: passed-in buffer filled with '\n'-terminated C string + * containing the current recovery pathname setting; + * return code is the size in bytes of the string + * On error: return code is zero or a negative errno value + */ static ssize_t write_recoverydir(struct file *file, char *buf, size_t size) { ssize_t rv; -- cgit v1.2.3-70-g09d2 From 3ada8b7e980dac7cc42937d42d90ee51b19204fe Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Tue, 6 Jan 2009 10:44:43 -0800 Subject: block: struct device - replace bus_id with dev_name(), dev_set_name() Cc: Jens Axboe Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman --- block/bsg.c | 6 +++--- block/genhd.c | 2 +- fs/partitions/check.c | 11 +++-------- 3 files changed, 7 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/block/bsg.c b/block/bsg.c index e73e50daf3d..d414bb5607e 100644 --- a/block/bsg.c +++ b/block/bsg.c @@ -42,7 +42,7 @@ struct bsg_device { int done_cmds; wait_queue_head_t wq_done; wait_queue_head_t wq_free; - char name[BUS_ID_SIZE]; + char name[20]; int max_queue; unsigned long flags; }; @@ -781,7 +781,7 @@ static struct bsg_device *bsg_add_device(struct inode *inode, mutex_lock(&bsg_mutex); hlist_add_head(&bd->dev_list, bsg_dev_idx_hash(iminor(inode))); - strncpy(bd->name, rq->bsg_dev.class_dev->bus_id, sizeof(bd->name) - 1); + strncpy(bd->name, dev_name(rq->bsg_dev.class_dev), sizeof(bd->name) - 1); dprintk("bound to <%s>, max queue %d\n", format_dev_t(buf, inode->i_rdev), bd->max_queue); @@ -992,7 +992,7 @@ int bsg_register_queue(struct request_queue *q, struct device *parent, if (name) devname = name; else - devname = parent->bus_id; + devname = dev_name(parent); /* * we need a proper transport to send commands, not a stacked device diff --git a/block/genhd.c b/block/genhd.c index d84a7df1e2a..397960cf26a 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -1084,7 +1084,7 @@ dev_t blk_lookup_devt(const char *name, int partno) struct gendisk *disk = dev_to_disk(dev); struct hd_struct *part; - if (strcmp(dev->bus_id, name)) + if (strcmp(dev_name(dev), name)) continue; part = disk_get_part(disk, partno); diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 6d5b213b8a9..5198ada6739 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -384,9 +384,9 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno, dname = dev_name(ddev); if (isdigit(dname[strlen(dname) - 1])) - snprintf(pdev->bus_id, BUS_ID_SIZE, "%sp%d", dname, partno); + dev_set_name(pdev, "%sp%d", dname, partno); else - snprintf(pdev->bus_id, BUS_ID_SIZE, "%s%d", dname, partno); + dev_set_name(pdev, "%s%d", dname, partno); device_initialize(pdev); pdev->class = &block_class; @@ -447,16 +447,11 @@ void register_disk(struct gendisk *disk) struct block_device *bdev; struct disk_part_iter piter; struct hd_struct *part; - char *s; int err; ddev->parent = disk->driverfs_dev; - strlcpy(ddev->bus_id, disk->disk_name, BUS_ID_SIZE); - /* ewww... some of these buggers have / in the name... */ - s = strchr(ddev->bus_id, '/'); - if (s) - *s = '!'; + dev_set_name(ddev, disk->disk_name); /* delay uevents, until we scanned partition table */ ddev->uevent_suppress = 1; -- cgit v1.2.3-70-g09d2 From 08fba69986e20c1c9e5fe2e6064d146cc4f42480 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 6 Jan 2009 14:38:53 -0800 Subject: mm: report the pagesize backing a VMA in /proc/pid/smaps It is useful to verify a hugepage-aware application is using the expected pagesizes for its memory regions. This patch creates an entry called KernelPageSize in /proc/pid/smaps that is the size of page used by the kernel to back a VMA. The entry is not called PageSize as it is possible the MMU uses a different size. This extension should not break any sensible parser that skips lines containing unrecognised information. Signed-off-by: Mel Gorman Acked-by: "KOSAKI Motohiro" Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/task_mmu.c | 6 ++++-- include/linux/hugetlb.h | 3 +++ mm/hugetlb.c | 16 ++++++++++++++++ 3 files changed, 23 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 3a8bdd7f575..41ef5f23e77 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -396,7 +396,8 @@ static int show_smap(struct seq_file *m, void *v) "Private_Clean: %8lu kB\n" "Private_Dirty: %8lu kB\n" "Referenced: %8lu kB\n" - "Swap: %8lu kB\n", + "Swap: %8lu kB\n" + "KernelPageSize: %8lu kB\n", (vma->vm_end - vma->vm_start) >> 10, mss.resident >> 10, (unsigned long)(mss.pss >> (10 + PSS_SHIFT)), @@ -405,7 +406,8 @@ static int show_smap(struct seq_file *m, void *v) mss.private_clean >> 10, mss.private_dirty >> 10, mss.referenced >> 10, - mss.swap >> 10); + mss.swap >> 10, + vma_kernel_pagesize(vma) >> 10); if (m->count < m->size) /* vma is copied successfully */ m->version = (vma != get_gate_vma(task)) ? vma->vm_start : 0; diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index e1c8afc002c..648e1e25979 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -233,6 +233,8 @@ static inline unsigned long huge_page_size(struct hstate *h) return (unsigned long)PAGE_SIZE << h->order; } +extern unsigned long vma_kernel_pagesize(struct vm_area_struct *vma); + static inline unsigned long huge_page_mask(struct hstate *h) { return h->mask; @@ -273,6 +275,7 @@ struct hstate {}; #define hstate_inode(i) NULL #define huge_page_size(h) PAGE_SIZE #define huge_page_mask(h) PAGE_MASK +#define vma_kernel_pagesize(v) PAGE_SIZE #define huge_page_order(h) 0 #define huge_page_shift(h) PAGE_SHIFT static inline unsigned int pages_per_huge_page(struct hstate *h) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 6058b53dcb8..5cb8bc7c80f 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -219,6 +219,22 @@ static pgoff_t vma_hugecache_offset(struct hstate *h, (vma->vm_pgoff >> huge_page_order(h)); } +/* + * Return the size of the pages allocated when backing a VMA. In the majority + * cases this will be same size as used by the page table entries. + */ +unsigned long vma_kernel_pagesize(struct vm_area_struct *vma) +{ + struct hstate *hstate; + + if (!is_vm_hugetlb_page(vma)) + return PAGE_SIZE; + + hstate = hstate_vma(vma); + + return 1UL << (hstate->order + PAGE_SHIFT); +} + /* * Flags for MAP_PRIVATE reservations. These are stored in the bottom * bits of the reservation map pointer, which are always clear due to -- cgit v1.2.3-70-g09d2 From 3340289ddf29ca75c3acfb3a6b72f234b2f74d5c Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 6 Jan 2009 14:38:54 -0800 Subject: mm: report the MMU pagesize in /proc/pid/smaps The KernelPageSize entry in /proc/pid/smaps is the pagesize used by the kernel to back a VMA. This matches the size used by the MMU in the majority of cases. However, one counter-example occurs on PPC64 kernels whereby a kernel using 64K as a base pagesize may still use 4K pages for the MMU on older processor. To distinguish, this patch reports MMUPageSize as the pagesize used by the MMU in /proc/pid/smaps. Signed-off-by: Mel Gorman Cc: "KOSAKI Motohiro" Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/include/asm/hugetlb.h | 6 ++++++ arch/powerpc/mm/hugetlbpage.c | 7 +++++++ fs/proc/task_mmu.c | 6 ++++-- include/linux/hugetlb.h | 3 +++ mm/hugetlb.c | 13 +++++++++++++ 5 files changed, 33 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/arch/powerpc/include/asm/hugetlb.h b/arch/powerpc/include/asm/hugetlb.h index 26f0d0ab27a..b1dafb6a974 100644 --- a/arch/powerpc/include/asm/hugetlb.h +++ b/arch/powerpc/include/asm/hugetlb.h @@ -17,6 +17,12 @@ void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep); +/* + * The version of vma_mmu_pagesize() in arch/powerpc/mm/hugetlbpage.c needs + * to override the version in mm/hugetlb.c + */ +#define vma_mmu_pagesize vma_mmu_pagesize + /* * If the arch doesn't supply something else, assume that hugepage * size aligned regions are ok without further preparation. diff --git a/arch/powerpc/mm/hugetlbpage.c b/arch/powerpc/mm/hugetlbpage.c index 201c7a5486c..9920d6a7cf2 100644 --- a/arch/powerpc/mm/hugetlbpage.c +++ b/arch/powerpc/mm/hugetlbpage.c @@ -512,6 +512,13 @@ unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr, return slice_get_unmapped_area(addr, len, flags, mmu_psize, 1, 0); } +unsigned long vma_mmu_pagesize(struct vm_area_struct *vma) +{ + unsigned int psize = get_slice_psize(vma->vm_mm, vma->vm_start); + + return 1UL << mmu_psize_to_shift(psize); +} + /* * Called by asm hashtable.S for doing lazy icache flush */ diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 41ef5f23e77..94063840832 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -397,7 +397,8 @@ static int show_smap(struct seq_file *m, void *v) "Private_Dirty: %8lu kB\n" "Referenced: %8lu kB\n" "Swap: %8lu kB\n" - "KernelPageSize: %8lu kB\n", + "KernelPageSize: %8lu kB\n" + "MMUPageSize: %8lu kB\n", (vma->vm_end - vma->vm_start) >> 10, mss.resident >> 10, (unsigned long)(mss.pss >> (10 + PSS_SHIFT)), @@ -407,7 +408,8 @@ static int show_smap(struct seq_file *m, void *v) mss.private_dirty >> 10, mss.referenced >> 10, mss.swap >> 10, - vma_kernel_pagesize(vma) >> 10); + vma_kernel_pagesize(vma) >> 10, + vma_mmu_pagesize(vma) >> 10); if (m->count < m->size) /* vma is copied successfully */ m->version = (vma != get_gate_vma(task)) ? vma->vm_start : 0; diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 648e1e25979..f1d2fba19ea 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -235,6 +235,8 @@ static inline unsigned long huge_page_size(struct hstate *h) extern unsigned long vma_kernel_pagesize(struct vm_area_struct *vma); +extern unsigned long vma_mmu_pagesize(struct vm_area_struct *vma); + static inline unsigned long huge_page_mask(struct hstate *h) { return h->mask; @@ -276,6 +278,7 @@ struct hstate {}; #define huge_page_size(h) PAGE_SIZE #define huge_page_mask(h) PAGE_MASK #define vma_kernel_pagesize(v) PAGE_SIZE +#define vma_mmu_pagesize(v) PAGE_SIZE #define huge_page_order(h) 0 #define huge_page_shift(h) PAGE_SHIFT static inline unsigned int pages_per_huge_page(struct hstate *h) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 5cb8bc7c80f..9595278b5ab 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -235,6 +235,19 @@ unsigned long vma_kernel_pagesize(struct vm_area_struct *vma) return 1UL << (hstate->order + PAGE_SHIFT); } +/* + * Return the page size being used by the MMU to back a VMA. In the majority + * of cases, the page size used by the kernel matches the MMU size. On + * architectures where it differs, an architecture-specific version of this + * function is required. + */ +#ifndef vma_mmu_pagesize +unsigned long vma_mmu_pagesize(struct vm_area_struct *vma) +{ + return vma_kernel_pagesize(vma); +} +#endif + /* * Flags for MAP_PRIVATE reservations. These are stored in the bottom * bits of the reservation map pointer, which are always clear due to -- cgit v1.2.3-70-g09d2 From 38c8e6180939e5619140b2e9e479cb26029ff8b1 Mon Sep 17 00:00:00 2001 From: Miquel van Smoorenburg Date: Tue, 6 Jan 2009 14:39:02 -0800 Subject: do_mpage_readpage(): don't submit lots of small bios on boundary While tracing I/O patterns with blktrace (a great tool) a few weeks ago I identified a minor issue in fs/mpage.c As the comment above mpage_readpages() says, a fs's get_block function will set BH_Boundary when it maps a block just before a block for which extra I/O is required. Since get_block() can map a range of pages, for all these pages the BH_Boundary flag will be set. But we only need to push what I/O we have accumulated at the last block of this range. This makes do_mpage_readpage() send out the largest possible bio instead of a bunch of page-sized ones in the BH_Boundary case. Signed-off-by: Miquel van Smoorenburg Cc: Nick Piggin Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/mpage.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/mpage.c b/fs/mpage.c index 552b80b3fac..46e977efd50 100644 --- a/fs/mpage.c +++ b/fs/mpage.c @@ -308,7 +308,10 @@ alloc_new: goto alloc_new; } - if (buffer_boundary(map_bh) || (first_hole != blocks_per_page)) + relative_block = block_in_file - *first_logical_block; + nblocks = map_bh->b_size >> blkbits; + if ((buffer_boundary(map_bh) && relative_block == nblocks) || + (first_hole != blocks_per_page)) bio = mpage_bio_submit(READ, bio); else *last_block_in_bio = blocks[blocks_per_page - 1]; -- cgit v1.2.3-70-g09d2 From ee53a891f47444c53318b98dac947ede963db400 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:39:12 -0800 Subject: mm: do_sync_mapping_range integrity fix Chris Mason notices do_sync_mapping_range didn't actually ask for data integrity writeout. Unfortunately, it is advertised as being usable for data integrity operations. This is a data integrity bug. Signed-off-by: Nick Piggin Cc: Chris Mason Cc: Dave Chinner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/sync.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/sync.c b/fs/sync.c index 0921d6d4b5e..ac02b56548b 100644 --- a/fs/sync.c +++ b/fs/sync.c @@ -295,7 +295,7 @@ int do_sync_mapping_range(struct address_space *mapping, loff_t offset, if (flags & SYNC_FILE_RANGE_WRITE) { ret = __filemap_fdatawrite_range(mapping, offset, endbyte, - WB_SYNC_NONE); + WB_SYNC_ALL); if (ret < 0) goto out; } -- cgit v1.2.3-70-g09d2 From 39f0dee2d8abe902617622b71f8f6f73985ec71c Mon Sep 17 00:00:00 2001 From: Franck Bui-Huu Date: Tue, 6 Jan 2009 14:39:21 -0800 Subject: do_mpage_readpage(): remove useless clear_buffer_mapped() call It is known that buffer_mapped() is false in this code path. Signed-off-by: Franck Bui-Huu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/mpage.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/mpage.c b/fs/mpage.c index 46e977efd50..16c3ef37eae 100644 --- a/fs/mpage.c +++ b/fs/mpage.c @@ -241,7 +241,6 @@ do_mpage_readpage(struct bio *bio, struct page *page, unsigned nr_pages, first_hole = page_block; page_block++; block_in_file++; - clear_buffer_mapped(map_bh); continue; } -- cgit v1.2.3-70-g09d2 From 3c1d43787b48c798f44dc32a6e6deb5ca2da3e68 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Tue, 6 Jan 2009 14:39:23 -0800 Subject: mm: remove GFP_HIGHUSER_PAGECACHE GFP_HIGHUSER_PAGECACHE is just an alias for GFP_HIGHUSER_MOVABLE, making that harder to track down: remove it, and its out-of-work brothers GFP_NOFS_PAGECACHE and GFP_USER_PAGECACHE. Since we're making that improvement to hotremove_migrate_alloc(), I think we can now also remove one of the "o"s from its comment. Signed-off-by: Hugh Dickins Acked-by: Mel Gorman Cc: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/inode.c | 4 ++-- include/linux/gfp.h | 6 ------ mm/memory_hotplug.c | 9 +++------ 3 files changed, 5 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index bd48e5e6d3e..a903860bc5a 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -166,7 +166,7 @@ struct inode *inode_init_always(struct super_block *sb, struct inode *inode) mapping->a_ops = &empty_aops; mapping->host = inode; mapping->flags = 0; - mapping_set_gfp_mask(mapping, GFP_HIGHUSER_PAGECACHE); + mapping_set_gfp_mask(mapping, GFP_HIGHUSER_MOVABLE); mapping->assoc_mapping = NULL; mapping->backing_dev_info = &default_backing_dev_info; mapping->writeback_index = 0; @@ -601,7 +601,7 @@ EXPORT_SYMBOL_GPL(inode_add_to_lists); * @sb: superblock * * Allocates a new inode for given superblock. The default gfp_mask - * for allocations related to inode->i_mapping is GFP_HIGHUSER_PAGECACHE. + * for allocations related to inode->i_mapping is GFP_HIGHUSER_MOVABLE. * If HIGHMEM pages are unsuitable or it is known that pages allocated * for the page cache are not reclaimable or migratable, * mapping_set_gfp_mask() must be called with suitable flags on the diff --git a/include/linux/gfp.h b/include/linux/gfp.h index e8003afeffb..dd20cd78faa 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -69,12 +69,6 @@ struct vm_area_struct; #define GFP_HIGHUSER_MOVABLE (__GFP_WAIT | __GFP_IO | __GFP_FS | \ __GFP_HARDWALL | __GFP_HIGHMEM | \ __GFP_MOVABLE) -#define GFP_NOFS_PAGECACHE (__GFP_WAIT | __GFP_IO | __GFP_MOVABLE) -#define GFP_USER_PAGECACHE (__GFP_WAIT | __GFP_IO | __GFP_FS | \ - __GFP_HARDWALL | __GFP_MOVABLE) -#define GFP_HIGHUSER_PAGECACHE (__GFP_WAIT | __GFP_IO | __GFP_FS | \ - __GFP_HARDWALL | __GFP_HIGHMEM | \ - __GFP_MOVABLE) #ifdef CONFIG_NUMA #define GFP_THISNODE (__GFP_THISNODE | __GFP_NOWARN | __GFP_NORETRY) diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 2ba38bc07b4..c083cf5fd6d 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -627,15 +627,12 @@ int scan_lru_pages(unsigned long start, unsigned long end) } static struct page * -hotremove_migrate_alloc(struct page *page, - unsigned long private, - int **x) +hotremove_migrate_alloc(struct page *page, unsigned long private, int **x) { - /* This should be improoooooved!! */ - return alloc_page(GFP_HIGHUSER_PAGECACHE); + /* This should be improooooved!! */ + return alloc_page(GFP_HIGHUSER_MOVABLE); } - #define NR_OFFLINE_AT_ONCE_PAGES (256) static int do_migrate_range(unsigned long start_pfn, unsigned long end_pfn) -- cgit v1.2.3-70-g09d2 From 0f64415d42760379753e6088787ce3fd3e069509 Mon Sep 17 00:00:00 2001 From: Dmitri Monakhov Date: Tue, 6 Jan 2009 14:40:04 -0800 Subject: fs: truncate blocks outside i_size after O_DIRECT write error In case of error extending write may have instantiated a few blocks outside i_size. We need to trim these blocks. We have to do it *regardless* to blocksize. At least ext2, ext3 and reiserfs interpret (i_size < biggest block) condition as error. Fsck will complain about wrong i_size. Then fsck will fix the error by changing i_size according to the biggest block. This is bad because this blocks contain garbage from previous write attempt. And result in data corruption. ####TESTCASE_BEGIN $touch /mnt/test/BIG_FILE ## at this moment /mnt/test/BIG_FILE size and blocks equal to zero open("/mnt/test/BIG_FILE", O_WRONLY|O_CREAT|O_DIRECT, 0666) = 3 write(3, "aaaaaaaaaaaa"..., 104857600) = -1 ENOSPC (No space left on device) ## size and block sould't be changed because write op failed. $stat /mnt/test/BIG_FILE File: `/mnt/test/BIG_FILE' Size: 0 Blocks: 110896 IO Block: 1024 regular empty file <<<<<<<<^^^^^^^^^^^^^^^^^^^^^^^^^^^^^file size is less than biggest block idx Device: fe07h/65031d Inode: 14 Links: 1 Access: (0644/-rw-r--r--) Uid: ( 0/ root) Gid: ( 0/ root) Access: 2007-01-24 20:03:38.000000000 +0300 Modify: 2007-01-24 20:03:38.000000000 +0300 Change: 2007-01-24 20:03:39.000000000 +0300 #fsck.ext3 -f /dev/VG/test e2fsck 1.39 (29-May-2006) Pass 1: Checking inodes, blocks, and sizes Inode 14, i_size is 0, should be 56556544. Fix? yes Pass 2: Checking directory structure .... #####TESTCASE_ENDdiff --git a/fs/direct-io.c b/fs/direct-io.c index af0558d..4e88bea 100644 [akpm@linux-foundation.org: use i_size_read()] Signed-off-by: Dmitri Monakhov Cc: Zach Brown Cc: Nick Piggin Cc: Badari Pulavarty Cc: Chris Mason Cc: Dave Chinner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/direct-io.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'fs') diff --git a/fs/direct-io.c b/fs/direct-io.c index af0558dbe8b..b6d43908ff7 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -1209,6 +1209,19 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, retval = direct_io_worker(rw, iocb, inode, iov, offset, nr_segs, blkbits, get_block, end_io, dio); + /* + * In case of error extending write may have instantiated a few + * blocks outside i_size. Trim these off again for DIO_LOCKING. + * NOTE: DIO_NO_LOCK/DIO_OWN_LOCK callers have to handle this by + * it's own meaner. + */ + if (unlikely(retval < 0 && (rw & WRITE))) { + loff_t isize = i_size_read(inode); + + if (end > isize && dio_lock_type == DIO_LOCKING) + vmtruncate(inode, isize); + } + if (rw == READ && dio_lock_type == DIO_LOCKING) release_i_mutex = 0; -- cgit v1.2.3-70-g09d2 From 91bf189c3a766927694ce9de7d545e96b23f20fc Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Tue, 6 Jan 2009 14:40:14 -0800 Subject: hugetlb: unsigned ret cannot be negative unsigned long ret cannot be negative, but ret can get -EFAULT. Signed-off-by: Roel Kluin Cc: Hugh Dickins Cc: Christoph Lameter Cc: Adam Litke Cc: David Gibson Cc: Ken Chen Cc: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hugetlbfs/inode.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 0ab0c6f5f43..6903d37af03 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -252,6 +252,7 @@ static ssize_t hugetlbfs_read(struct file *filp, char __user *buf, for (;;) { struct page *page; unsigned long nr, ret; + int ra; /* nr is the maximum number of bytes to copy from this page */ nr = huge_page_size(h); @@ -274,16 +275,19 @@ static ssize_t hugetlbfs_read(struct file *filp, char __user *buf, */ ret = len < nr ? len : nr; if (clear_user(buf, ret)) - ret = -EFAULT; + ra = -EFAULT; + else + ra = 0; } else { /* * We have the page, copy it to user space buffer. */ - ret = hugetlbfs_read_actor(page, offset, buf, len, nr); + ra = hugetlbfs_read_actor(page, offset, buf, len, nr); + ret = ra; } - if (ret < 0) { + if (ra < 0) { if (retval == 0) - retval = ret; + retval = ra; if (page) page_cache_release(page); goto out; -- cgit v1.2.3-70-g09d2 From 69e9930993cfd70d82c8d9dd96fc3a88854d06fc Mon Sep 17 00:00:00 2001 From: Franck Bui-Huu Date: Tue, 6 Jan 2009 14:40:19 -0800 Subject: block_write_begin(): remove useless goto Signed-off-by: Franck Bui-Huu Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index a13f09b696f..c26da785938 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -2022,7 +2022,6 @@ int block_write_begin(struct file *file, struct address_space *mapping, if (pos + len > inode->i_size) vmtruncate(inode, inode->i_size); } - goto out; } out: -- cgit v1.2.3-70-g09d2 From e8ea1759138d4279869f52bfb7dca8f02f8ccfe5 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 6 Jan 2009 14:40:23 -0800 Subject: UBIFS: do not use WB_SYNC_HOLD WB_SYNC_HOLD is going to be zapped so we should not use it. Use %WB_SYNC_NONE instead. Here is what akpm said: "I think I'll just switch that to WB_SYNC_NONE. The `wait==0' mode is just an advisory thing to help the fs shove lots of data into the queues. If some gets missed then it'll be picked up on the second ->sync_fs call, with wait==1." Thanks to Randy Dunlap for catching this. Signed-off-by: Artem Bityutskiy Cc: Nick Piggin Cc: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ubifs/super.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 0d7564b95f8..89556ee7251 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -432,12 +432,19 @@ static int ubifs_sync_fs(struct super_block *sb, int wait) int i, err; struct ubifs_info *c = sb->s_fs_info; struct writeback_control wbc = { - .sync_mode = wait ? WB_SYNC_ALL : WB_SYNC_HOLD, + .sync_mode = wait ? WB_SYNC_ALL : WB_SYNC_NONE, .range_start = 0, .range_end = LLONG_MAX, .nr_to_write = LONG_MAX, }; + /* + * Note by akpm about WB_SYNC_NONE used above: zero @wait is just an + * advisory thing to help the file system shove lots of data into the + * queues. If some gets missed then it'll be picked up on the second + * '->sync_fs()' call, with non-zero @wait. + */ + if (sb->s_flags & MS_RDONLY) return 0; -- cgit v1.2.3-70-g09d2 From 4f5a99d64c17470a784a6c68064207d82e3e74a5 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:40:25 -0800 Subject: fs: remove WB_SYNC_HOLD Remove WB_SYNC_HOLD. The primary motiviation is the design of my anti-starvation code for fsync. It requires taking an inode lock over the sync operation, so we could run into lock ordering problems with multiple inodes. It is possible to take a single global lock to solve the ordering problem, but then that would prevent a future nice implementation of "sync multiple inodes" based on lock order via inode address. Seems like a backward step to remove this, but actually it is busted anyway: we can't use the inode lists for data integrity wait: an inode can be taken off the dirty lists but still be under writeback. In order to satisfy data integrity semantics, we should wait for it to finish writeback, but if we only search the dirty lists, we'll miss it. It would be possible to have a "writeback" list, for sys_sync, I suppose. But why complicate things by prematurely optimise? For unmounting, we could avoid the "livelock avoidance" code, which would be easier, but again premature IMO. Fixing the existing data integrity problem will come next. Signed-off-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fs-writeback.c | 12 ++---------- include/linux/writeback.h | 1 - 2 files changed, 2 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index d0ff0b8cf30..d99601af9e4 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -421,9 +421,6 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc) * If we're a pdlfush thread, then implement pdflush collision avoidance * against the entire list. * - * WB_SYNC_HOLD is a hack for sys_sync(): reattach the inode to sb->s_dirty so - * that it can be located for waiting on in __writeback_single_inode(). - * * If `bdi' is non-zero then we're being asked to writeback a specific queue. * This function assumes that the blockdev superblock's inodes are backed by * a variety of queues, so all inodes are searched. For other superblocks, @@ -499,10 +496,6 @@ void generic_sync_sb_inodes(struct super_block *sb, __iget(inode); pages_skipped = wbc->pages_skipped; __writeback_single_inode(inode, wbc); - if (wbc->sync_mode == WB_SYNC_HOLD) { - inode->dirtied_when = jiffies; - list_move(&inode->i_list, &sb->s_dirty); - } if (current_is_pdflush()) writeback_release(bdi); if (wbc->pages_skipped != pages_skipped) { @@ -588,8 +581,7 @@ restart: /* * writeback and wait upon the filesystem's dirty inodes. The caller will - * do this in two passes - one to write, and one to wait. WB_SYNC_HOLD is - * used to park the written inodes on sb->s_dirty for the wait pass. + * do this in two passes - one to write, and one to wait. * * A finite limit is set on the number of pages which will be written. * To prevent infinite livelock of sys_sync(). @@ -600,7 +592,7 @@ restart: void sync_inodes_sb(struct super_block *sb, int wait) { struct writeback_control wbc = { - .sync_mode = wait ? WB_SYNC_ALL : WB_SYNC_HOLD, + .sync_mode = wait ? WB_SYNC_ALL : WB_SYNC_NONE, .range_start = 0, .range_end = LLONG_MAX, }; diff --git a/include/linux/writeback.h b/include/linux/writeback.h index bb28c975c1d..7300ecdc480 100644 --- a/include/linux/writeback.h +++ b/include/linux/writeback.h @@ -30,7 +30,6 @@ static inline int task_is_pdflush(struct task_struct *task) enum writeback_sync_modes { WB_SYNC_NONE, /* Don't wait on anything */ WB_SYNC_ALL, /* Wait on every mapping */ - WB_SYNC_HOLD, /* Hold the inode on sb_dirty for sys_sync() */ }; /* -- cgit v1.2.3-70-g09d2 From 38f21977663126fef53f5585e7f1653d8ebe55c4 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:40:25 -0800 Subject: fs: sync_sb_inodes fix Fix data integrity semantics required by sys_sync, by iterating over all inodes and waiting for any writeback pages after the initial writeout. Comments explain the exact problem. Signed-off-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fs-writeback.c | 60 ++++++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 53 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index d99601af9e4..a9ee474f969 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -440,6 +440,7 @@ void generic_sync_sb_inodes(struct super_block *sb, struct writeback_control *wbc) { const unsigned long start = jiffies; /* livelock avoidance */ + int sync = wbc->sync_mode == WB_SYNC_ALL; spin_lock(&inode_lock); if (!wbc->for_kupdate || list_empty(&sb->s_io)) @@ -516,7 +517,49 @@ void generic_sync_sb_inodes(struct super_block *sb, if (!list_empty(&sb->s_more_io)) wbc->more_io = 1; } - spin_unlock(&inode_lock); + + if (sync) { + struct inode *inode, *old_inode = NULL; + + /* + * Data integrity sync. Must wait for all pages under writeback, + * because there may have been pages dirtied before our sync + * call, but which had writeout started before we write it out. + * In which case, the inode may not be on the dirty list, but + * we still have to wait for that writeout. + */ + list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { + struct address_space *mapping; + + if (inode->i_state & (I_FREEING|I_WILL_FREE)) + continue; + mapping = inode->i_mapping; + if (mapping->nrpages == 0) + continue; + __iget(inode); + spin_unlock(&inode_lock); + /* + * We hold a reference to 'inode' so it couldn't have + * been removed from s_inodes list while we dropped the + * inode_lock. We cannot iput the inode now as we can + * be holding the last reference and we cannot iput it + * under inode_lock. So we keep the reference and iput + * it later. + */ + iput(old_inode); + old_inode = inode; + + filemap_fdatawait(mapping); + + cond_resched(); + + spin_lock(&inode_lock); + } + spin_unlock(&inode_lock); + iput(old_inode); + } else + spin_unlock(&inode_lock); + return; /* Leave any unwritten inodes on s_io */ } EXPORT_SYMBOL_GPL(generic_sync_sb_inodes); @@ -596,13 +639,16 @@ void sync_inodes_sb(struct super_block *sb, int wait) .range_start = 0, .range_end = LLONG_MAX, }; - unsigned long nr_dirty = global_page_state(NR_FILE_DIRTY); - unsigned long nr_unstable = global_page_state(NR_UNSTABLE_NFS); - wbc.nr_to_write = nr_dirty + nr_unstable + - (inodes_stat.nr_inodes - inodes_stat.nr_unused) + - nr_dirty + nr_unstable; - wbc.nr_to_write += wbc.nr_to_write / 2; /* Bit more for luck */ + if (!wait) { + unsigned long nr_dirty = global_page_state(NR_FILE_DIRTY); + unsigned long nr_unstable = global_page_state(NR_UNSTABLE_NFS); + + wbc.nr_to_write = nr_dirty + nr_unstable + + (inodes_stat.nr_inodes - inodes_stat.nr_unused); + } else + wbc.nr_to_write = LONG_MAX; /* doesn't actually matter */ + sync_sb_inodes(sb, &wbc); } -- cgit v1.2.3-70-g09d2 From 856bf4d717feb8c55d4e2f817b71ebb70cfbc67b Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:40:26 -0800 Subject: fs: sys_sync fix s_syncing livelock avoidance was breaking data integrity guarantee of sys_sync, by allowing sys_sync to skip writing or waiting for superblocks if there is a concurrent sys_sync happening. This livelock avoidance is much less important now that we don't have the get_super_to_sync() call after every sb that we sync. This was replaced by __put_super_and_need_restart. Signed-off-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fs-writeback.c | 20 +------------------- include/linux/fs.h | 1 - 2 files changed, 1 insertion(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index a9ee474f969..e5eaa62fd17 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -652,18 +652,6 @@ void sync_inodes_sb(struct super_block *sb, int wait) sync_sb_inodes(sb, &wbc); } -/* - * Rather lame livelock avoidance. - */ -static void set_sb_syncing(int val) -{ - struct super_block *sb; - spin_lock(&sb_lock); - list_for_each_entry_reverse(sb, &super_blocks, s_list) - sb->s_syncing = val; - spin_unlock(&sb_lock); -} - /** * sync_inodes - writes all inodes to disk * @wait: wait for completion @@ -690,9 +678,6 @@ static void __sync_inodes(int wait) spin_lock(&sb_lock); restart: list_for_each_entry(sb, &super_blocks, s_list) { - if (sb->s_syncing) - continue; - sb->s_syncing = 1; sb->s_count++; spin_unlock(&sb_lock); down_read(&sb->s_umount); @@ -710,13 +695,10 @@ restart: void sync_inodes(int wait) { - set_sb_syncing(0); __sync_inodes(0); - if (wait) { - set_sb_syncing(0); + if (wait) __sync_inodes(1); - } } /** diff --git a/include/linux/fs.h b/include/linux/fs.h index fb59673c60b..d7eba77f666 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1133,7 +1133,6 @@ struct super_block { struct rw_semaphore s_umount; struct mutex s_lock; int s_count; - int s_syncing; int s_need_sync_fs; atomic_t s_active; #ifdef CONFIG_SECURITY -- cgit v1.2.3-70-g09d2 From eaccbfa564e48c87626594511f42dc8c0ad2daae Mon Sep 17 00:00:00 2001 From: "Luiz Fernando N. Capitulino" Date: Tue, 6 Jan 2009 14:40:44 -0800 Subject: fs/exec.c:__bprm_mm_init(): clean up error handling Untangle the error unwinding in this function, saving a test of local variable `vma'. Signed-off-by: Luiz Fernando N. Capitulino Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/exec.c | 20 ++++++-------------- 1 file changed, 6 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/exec.c b/fs/exec.c index 9c33f542dc7..6ca4e3d22a3 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -232,13 +232,13 @@ static void flush_arg_page(struct linux_binprm *bprm, unsigned long pos, static int __bprm_mm_init(struct linux_binprm *bprm) { - int err = -ENOMEM; + int err; struct vm_area_struct *vma = NULL; struct mm_struct *mm = bprm->mm; bprm->vma = vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); if (!vma) - goto err; + return -ENOMEM; down_write(&mm->mmap_sem); vma->vm_mm = mm; @@ -251,28 +251,20 @@ static int __bprm_mm_init(struct linux_binprm *bprm) */ vma->vm_end = STACK_TOP_MAX; vma->vm_start = vma->vm_end - PAGE_SIZE; - vma->vm_flags = VM_STACK_FLAGS; vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); err = insert_vm_struct(mm, vma); - if (err) { - up_write(&mm->mmap_sem); + if (err) goto err; - } mm->stack_vm = mm->total_vm = 1; up_write(&mm->mmap_sem); - bprm->p = vma->vm_end - sizeof(void *); - return 0; - err: - if (vma) { - bprm->vma = NULL; - kmem_cache_free(vm_area_cachep, vma); - } - + up_write(&mm->mmap_sem); + bprm->vma = NULL; + kmem_cache_free(vm_area_cachep, vma); return err; } -- cgit v1.2.3-70-g09d2 From 67ec7d3ab779ad9001ef57a6b4cfdf80ac9f9acc Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 6 Jan 2009 14:40:57 -0800 Subject: fs: use menuconfig to control the Misc. filesystems menu Have one option to control Miscellaneous filesystems. This makes it easy to disable all of them at one time. Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/Kconfig | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/Kconfig b/fs/Kconfig index f9b6e2979aa..32883589ee5 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -721,7 +721,20 @@ config CONFIGFS_FS endmenu -menu "Miscellaneous filesystems" +menuconfig MISC_FILESYSTEMS + bool "Miscellaneous filesystems" + default y + ---help--- + Say Y here to get to see options for various miscellaneous + filesystems, such as filesystems that came from other + operating systems. + + This option alone does not add any kernel code. + + If you say N, all options in this submenu will be skipped and + disabled; if unsure, say Y here. + +if MISC_FILESYSTEMS config ADFS_FS tristate "ADFS file system support (EXPERIMENTAL)" @@ -1091,7 +1104,7 @@ config UFS_DEBUG Y here. This will result in _many_ additional debugging messages to be written to the system log. -endmenu +endif # MISC_FILESYSTEMS menuconfig NETWORK_FILESYSTEMS bool "Network File Systems" -- cgit v1.2.3-70-g09d2 From 5f820f648c92a5ecc771a96b3c29aa6e90013bba Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 6 Jan 2009 14:40:59 -0800 Subject: poll: allow f_op->poll to sleep f_op->poll is the only vfs operation which is not allowed to sleep. It's because poll and select implementation used task state to synchronize against wake ups, which doesn't have to be the case anymore as wait/wake interface can now use custom wake up functions. The non-sleep restriction can be a bit tricky because ->poll is not called from an atomic context and the result of accidentally sleeping in ->poll only shows up as temporary busy looping when the timing is right or rather wrong. This patch converts poll/select to use custom wake up function and use separate triggered variable to synchronize against wake up events. The only added overhead is an extra function call during wake up and negligible. This patch removes the one non-sleep exception from vfs locking rules and is beneficial to userland filesystem implementations like FUSE, 9p or peculiar fs like spufs as it's very difficult for those to implement non-sleeping poll method. While at it, make the following cosmetic changes to make poll.h and select.c checkpatch friendly. * s/type * symbol/type *symbol/ : three places in poll.h * remove blank line before EXPORT_SYMBOL() : two places in select.c Oleg: spotted missing barrier in poll_schedule_timeout() Davide: spotted missing write barrier in pollwake() Signed-off-by: Tejun Heo Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Ingo Molnar Cc: Christoph Hellwig Signed-off-by: Miklos Szeredi Cc: Davide Libenzi Cc: Brad Boyer Cc: Al Viro Cc: Roland McGrath Cc: Mauro Carvalho Chehab Signed-off-by: Andrew Morton Cc: Davide Libenzi Cc: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/Locking | 2 +- drivers/media/video/v4l1-compat.c | 4 +-- fs/select.c | 76 +++++++++++++++++++++++++++++++-------- include/linux/poll.h | 15 ++++++-- 4 files changed, 76 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index ccec5539438..cfbfa15a46b 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -397,7 +397,7 @@ prototypes: }; locking rules: - All except ->poll() may block. + All may block. BKL llseek: no (see below) read: no diff --git a/drivers/media/video/v4l1-compat.c b/drivers/media/video/v4l1-compat.c index d450cab20be..b617bf05e2d 100644 --- a/drivers/media/video/v4l1-compat.c +++ b/drivers/media/video/v4l1-compat.c @@ -203,7 +203,6 @@ static int poll_one(struct file *file, struct poll_wqueues *pwq) table = &pwq->pt; for (;;) { int mask; - set_current_state(TASK_INTERRUPTIBLE); mask = file->f_op->poll(file, table); if (mask & POLLIN) break; @@ -212,9 +211,8 @@ static int poll_one(struct file *file, struct poll_wqueues *pwq) retval = -ERESTARTSYS; break; } - schedule(); + poll_schedule(pwq, TASK_INTERRUPTIBLE); } - set_current_state(TASK_RUNNING); poll_freewait(pwq); return retval; } diff --git a/fs/select.c b/fs/select.c index 87df51eadcf..08b91beed80 100644 --- a/fs/select.c +++ b/fs/select.c @@ -109,11 +109,11 @@ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, void poll_initwait(struct poll_wqueues *pwq) { init_poll_funcptr(&pwq->pt, __pollwait); + pwq->polling_task = current; pwq->error = 0; pwq->table = NULL; pwq->inline_index = 0; } - EXPORT_SYMBOL(poll_initwait); static void free_poll_entry(struct poll_table_entry *entry) @@ -142,12 +142,10 @@ void poll_freewait(struct poll_wqueues *pwq) free_page((unsigned long) old); } } - EXPORT_SYMBOL(poll_freewait); -static struct poll_table_entry *poll_get_entry(poll_table *_p) +static struct poll_table_entry *poll_get_entry(struct poll_wqueues *p) { - struct poll_wqueues *p = container_of(_p, struct poll_wqueues, pt); struct poll_table_page *table = p->table; if (p->inline_index < N_INLINE_POLL_ENTRIES) @@ -159,7 +157,6 @@ static struct poll_table_entry *poll_get_entry(poll_table *_p) new_table = (struct poll_table_page *) __get_free_page(GFP_KERNEL); if (!new_table) { p->error = -ENOMEM; - __set_current_state(TASK_RUNNING); return NULL; } new_table->entry = new_table->entries; @@ -171,20 +168,75 @@ static struct poll_table_entry *poll_get_entry(poll_table *_p) return table->entry++; } +static int pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key) +{ + struct poll_wqueues *pwq = wait->private; + DECLARE_WAITQUEUE(dummy_wait, pwq->polling_task); + + /* + * Although this function is called under waitqueue lock, LOCK + * doesn't imply write barrier and the users expect write + * barrier semantics on wakeup functions. The following + * smp_wmb() is equivalent to smp_wmb() in try_to_wake_up() + * and is paired with set_mb() in poll_schedule_timeout. + */ + smp_wmb(); + pwq->triggered = 1; + + /* + * Perform the default wake up operation using a dummy + * waitqueue. + * + * TODO: This is hacky but there currently is no interface to + * pass in @sync. @sync is scheduled to be removed and once + * that happens, wake_up_process() can be used directly. + */ + return default_wake_function(&dummy_wait, mode, sync, key); +} + /* Add a new entry */ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p) { - struct poll_table_entry *entry = poll_get_entry(p); + struct poll_wqueues *pwq = container_of(p, struct poll_wqueues, pt); + struct poll_table_entry *entry = poll_get_entry(pwq); if (!entry) return; get_file(filp); entry->filp = filp; entry->wait_address = wait_address; - init_waitqueue_entry(&entry->wait, current); + init_waitqueue_func_entry(&entry->wait, pollwake); + entry->wait.private = pwq; add_wait_queue(wait_address, &entry->wait); } +int poll_schedule_timeout(struct poll_wqueues *pwq, int state, + ktime_t *expires, unsigned long slack) +{ + int rc = -EINTR; + + set_current_state(state); + if (!pwq->triggered) + rc = schedule_hrtimeout_range(expires, slack, HRTIMER_MODE_ABS); + __set_current_state(TASK_RUNNING); + + /* + * Prepare for the next iteration. + * + * The following set_mb() serves two purposes. First, it's + * the counterpart rmb of the wmb in pollwake() such that data + * written before wake up is always visible after wake up. + * Second, the full barrier guarantees that triggered clearing + * doesn't pass event check of the next iteration. Note that + * this problem doesn't exist for the first iteration as + * add_wait_queue() has full barrier semantics. + */ + set_mb(pwq->triggered, 0); + + return rc; +} +EXPORT_SYMBOL(poll_schedule_timeout); + /** * poll_select_set_timeout - helper function to setup the timeout value * @to: pointer to timespec variable for the final timeout @@ -340,8 +392,6 @@ int do_select(int n, fd_set_bits *fds, struct timespec *end_time) for (;;) { unsigned long *rinp, *routp, *rexp, *inp, *outp, *exp; - set_current_state(TASK_INTERRUPTIBLE); - inp = fds->in; outp = fds->out; exp = fds->ex; rinp = fds->res_in; routp = fds->res_out; rexp = fds->res_ex; @@ -411,10 +461,10 @@ int do_select(int n, fd_set_bits *fds, struct timespec *end_time) to = &expire; } - if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) + if (!poll_schedule_timeout(&table, TASK_INTERRUPTIBLE, + to, slack)) timed_out = 1; } - __set_current_state(TASK_RUNNING); poll_freewait(&table); @@ -666,7 +716,6 @@ static int do_poll(unsigned int nfds, struct poll_list *list, for (;;) { struct poll_list *walk; - set_current_state(TASK_INTERRUPTIBLE); for (walk = list; walk != NULL; walk = walk->next) { struct pollfd * pfd, * pfd_end; @@ -709,10 +758,9 @@ static int do_poll(unsigned int nfds, struct poll_list *list, to = &expire; } - if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) + if (!poll_schedule_timeout(wait, TASK_INTERRUPTIBLE, to, slack)) timed_out = 1; } - __set_current_state(TASK_RUNNING); return count; } diff --git a/include/linux/poll.h b/include/linux/poll.h index badd98ab06f..8c24ef8d997 100644 --- a/include/linux/poll.h +++ b/include/linux/poll.h @@ -46,9 +46,9 @@ static inline void init_poll_funcptr(poll_table *pt, poll_queue_proc qproc) } struct poll_table_entry { - struct file * filp; + struct file *filp; wait_queue_t wait; - wait_queue_head_t * wait_address; + wait_queue_head_t *wait_address; }; /* @@ -56,7 +56,9 @@ struct poll_table_entry { */ struct poll_wqueues { poll_table pt; - struct poll_table_page * table; + struct poll_table_page *table; + struct task_struct *polling_task; + int triggered; int error; int inline_index; struct poll_table_entry inline_entries[N_INLINE_POLL_ENTRIES]; @@ -64,6 +66,13 @@ struct poll_wqueues { extern void poll_initwait(struct poll_wqueues *pwq); extern void poll_freewait(struct poll_wqueues *pwq); +extern int poll_schedule_timeout(struct poll_wqueues *pwq, int state, + ktime_t *expires, unsigned long slack); + +static inline int poll_schedule(struct poll_wqueues *pwq, int state) +{ + return poll_schedule_timeout(pwq, state, NULL, 0); +} /* * Scaleable version of the fd_set. -- cgit v1.2.3-70-g09d2 From 179f7ebff6be45738c6e2fa68c8d2cc5c2c6308e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 6 Jan 2009 14:41:04 -0800 Subject: percpu_counter: FBC_BATCH should be a variable For NR_CPUS >= 16 values, FBC_BATCH is 2*NR_CPUS Considering more and more distros are using high NR_CPUS values, it makes sense to use a more sensible value for FBC_BATCH, and get rid of NR_CPUS. A sensible value is 2*num_online_cpus(), with a minimum value of 32 (This minimum value helps branch prediction in __percpu_counter_add()) We already have a hotcpu notifier, so we can adjust FBC_BATCH dynamically. We rename FBC_BATCH to percpu_counter_batch since its not a constant anymore. Signed-off-by: Eric Dumazet Acked-by: David S. Miller Acked-by: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext4/ext4.h | 6 +++--- fs/ext4/inode.c | 2 +- include/linux/percpu_counter.h | 8 ++------ lib/percpu_counter.c | 18 ++++++++++++++---- 4 files changed, 20 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index b0537c82702..6c46c648430 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1225,11 +1225,11 @@ do { \ } while (0) #ifdef CONFIG_SMP -/* Each CPU can accumulate FBC_BATCH blocks in their local +/* Each CPU can accumulate percpu_counter_batch blocks in their local * counters. So we need to make sure we have free blocks more - * than FBC_BATCH * nr_cpu_ids. Also add a window of 4 times. + * than percpu_counter_batch * nr_cpu_ids. Also add a window of 4 times. */ -#define EXT4_FREEBLOCKS_WATERMARK (4 * (FBC_BATCH * nr_cpu_ids)) +#define EXT4_FREEBLOCKS_WATERMARK (4 * (percpu_counter_batch * nr_cpu_ids)) #else #define EXT4_FREEBLOCKS_WATERMARK 0 #endif diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 6702a49992a..98d3fe7057e 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2498,7 +2498,7 @@ static int ext4_nonda_switch(struct super_block *sb) /* * switch to non delalloc mode if we are running low * on free block. The free block accounting via percpu - * counters can get slightly wrong with FBC_BATCH getting + * counters can get slightly wrong with percpu_counter_batch getting * accumulated on each CPU without updating global counters * Delalloc need an accurate free block accounting. So switch * to non delalloc when we are near to error range. diff --git a/include/linux/percpu_counter.h b/include/linux/percpu_counter.h index 9007ccdfc11..99de7a31bab 100644 --- a/include/linux/percpu_counter.h +++ b/include/linux/percpu_counter.h @@ -24,11 +24,7 @@ struct percpu_counter { s32 *counters; }; -#if NR_CPUS >= 16 -#define FBC_BATCH (NR_CPUS*2) -#else -#define FBC_BATCH (NR_CPUS*4) -#endif +extern int percpu_counter_batch; int percpu_counter_init(struct percpu_counter *fbc, s64 amount); int percpu_counter_init_irq(struct percpu_counter *fbc, s64 amount); @@ -39,7 +35,7 @@ s64 __percpu_counter_sum(struct percpu_counter *fbc); static inline void percpu_counter_add(struct percpu_counter *fbc, s64 amount) { - __percpu_counter_add(fbc, amount, FBC_BATCH); + __percpu_counter_add(fbc, amount, percpu_counter_batch); } static inline s64 percpu_counter_sum_positive(struct percpu_counter *fbc) diff --git a/lib/percpu_counter.c b/lib/percpu_counter.c index b255b939bc1..a60bd804609 100644 --- a/lib/percpu_counter.c +++ b/lib/percpu_counter.c @@ -9,10 +9,8 @@ #include #include -#ifdef CONFIG_HOTPLUG_CPU static LIST_HEAD(percpu_counters); static DEFINE_MUTEX(percpu_counters_lock); -#endif void percpu_counter_set(struct percpu_counter *fbc, s64 amount) { @@ -111,13 +109,24 @@ void percpu_counter_destroy(struct percpu_counter *fbc) } EXPORT_SYMBOL(percpu_counter_destroy); -#ifdef CONFIG_HOTPLUG_CPU +int percpu_counter_batch __read_mostly = 32; +EXPORT_SYMBOL(percpu_counter_batch); + +static void compute_batch_value(void) +{ + int nr = num_online_cpus(); + + percpu_counter_batch = max(32, nr*2); +} + static int __cpuinit percpu_counter_hotcpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) { +#ifdef CONFIG_HOTPLUG_CPU unsigned int cpu; struct percpu_counter *fbc; + compute_batch_value(); if (action != CPU_DEAD) return NOTIFY_OK; @@ -134,13 +143,14 @@ static int __cpuinit percpu_counter_hotcpu_callback(struct notifier_block *nb, spin_unlock_irqrestore(&fbc->lock, flags); } mutex_unlock(&percpu_counters_lock); +#endif return NOTIFY_OK; } static int __init percpu_counter_startup(void) { + compute_batch_value(); hotcpu_notifier(percpu_counter_hotcpu_callback, 0); return 0; } module_init(percpu_counter_startup); -#endif -- cgit v1.2.3-70-g09d2 From 8c4018884a49eb2c6c7ca90804f331b12983561c Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Tue, 6 Jan 2009 14:41:08 -0800 Subject: fs: fix name overwrite in __register_chrdev_region() It's possible to register a chrdev with a name size exactly the same as was allocated in structure. It seems it was not intended behaviour. At least chrdev_show does not like it. Signed-off-by: Cyrill Gorcunov Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/char_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/char_dev.c b/fs/char_dev.c index 700697a7261..38f71222a55 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -120,7 +120,7 @@ __register_chrdev_region(unsigned int major, unsigned int baseminor, cd->major = major; cd->baseminor = baseminor; cd->minorct = minorct; - strncpy(cd->name,name, 64); + strlcpy(cd->name, name, sizeof(cd->name)); i = major_to_index(major); -- cgit v1.2.3-70-g09d2 From ca8a5bd28226d62c045e0e55b42b9d10b146c205 Mon Sep 17 00:00:00 2001 From: Gerd Hoffmann Date: Tue, 6 Jan 2009 14:41:09 -0800 Subject: add missing accounting calls to compat_sys_{readv,writev} Signed-off-by: Gerd Hoffmann Cc: Jay Lan Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/compat.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/compat.c b/fs/compat.c index d1ece79b641..30f2faa22f5 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1187,6 +1187,9 @@ compat_sys_readv(unsigned long fd, const struct compat_iovec __user *vec, unsign ret = compat_do_readv_writev(READ, file, vec, vlen, &file->f_pos); out: + if (ret > 0) + add_rchar(current, ret); + inc_syscr(current); fput(file); return ret; } @@ -1210,6 +1213,9 @@ compat_sys_writev(unsigned long fd, const struct compat_iovec __user *vec, unsig ret = compat_do_readv_writev(WRITE, file, vec, vlen, &file->f_pos); out: + if (ret > 0) + add_wchar(current, ret); + inc_syscw(current); fput(file); return ret; } -- cgit v1.2.3-70-g09d2 From 350eaf791bebccb9ad5999351f3e328319545f03 Mon Sep 17 00:00:00 2001 From: Tetsuo Handa Date: Tue, 6 Jan 2009 14:41:11 -0800 Subject: do_coredump(): check return from argv_split() do_coredump() accesses helper_argv[0] without checking helper_argv != NULL. This can happen if page allocation failed. Signed-off-by: Tetsuo Handa Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/exec.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/exec.c b/fs/exec.c index 6ca4e3d22a3..6b09d6fa4f7 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1770,6 +1770,11 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) if (ispipe) { helper_argv = argv_split(GFP_KERNEL, corename+1, &helper_argc); + if (!helper_argv) { + printk(KERN_WARNING "%s failed to allocate memory\n", + __func__); + goto fail_unlock; + } /* Terminate the string before the first option */ delimit = strchr(corename, ' '); if (delimit) -- cgit v1.2.3-70-g09d2 From 0bc02f3fa433a98631a932e77c4b1f873da35aee Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 6 Jan 2009 14:41:13 -0800 Subject: fs/inode: fix kernel-doc notation Fix kernel-doc notation: Warning(linux-2.6.28-git3//fs/inode.c:120): No description found for parameter 'sb' Warning(linux-2.6.28-git3//fs/inode.c:120): No description found for parameter 'inode' Warning(linux-2.6.28-git3//fs/inode.c:588): No description found for parameter 'sb' Warning(linux-2.6.28-git3//fs/inode.c:588): No description found for parameter 'inode' Signed-off-by: Randy Dunlap Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/inode.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index a903860bc5a..7a6e8c2ff7b 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -110,8 +110,8 @@ static void wake_up_inode(struct inode *inode) /** * inode_init_always - perform inode structure intialisation - * @sb - superblock inode belongs to. - * @inode - inode to initialise + * @sb: superblock inode belongs to + * @inode: inode to initialise * * These are initializations that need to be done on every inode * allocation as the fields are not initialised by slab allocation. @@ -576,8 +576,8 @@ __inode_add_to_lists(struct super_block *sb, struct hlist_head *head, /** * inode_add_to_lists - add a new inode to relevant lists - * @sb - superblock inode belongs to. - * @inode - inode to mark in use + * @sb: superblock inode belongs to + * @inode: inode to mark in use * * When an inode is allocated it needs to be accounted for, added to the in use * list, the owning superblock and the inode hash. This needs to be done under -- cgit v1.2.3-70-g09d2 From 94e2959e7a6a4ef0969932c30349ce6f4469a3cf Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 6 Jan 2009 14:41:15 -0800 Subject: fs: fix function param name in kernel-doc Fix function parameter name in kernel-doc: Warning(linux-2.6.28-git5//fs/block_dev.c:1272): No description found for parameter 'pathname' Warning(linux-2.6.28-git5//fs/block_dev.c:1272): Excess function parameter 'path' description in 'lookup_bdev' Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/block_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 349a26c1000..b957717e25a 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1262,7 +1262,7 @@ EXPORT_SYMBOL(ioctl_by_bdev); /** * lookup_bdev - lookup a struct block_device by name - * @path: special file representing the block device + * @pathname: special file representing the block device * * Get a reference to the blockdevice at @pathname in the current * namespace if possible and return it. Return ERR_PTR(error) -- cgit v1.2.3-70-g09d2 From 87113e806a9ee48c6c989513ef3e9c1d31e06ac4 Mon Sep 17 00:00:00 2001 From: Qinghuang Feng Date: Tue, 6 Jan 2009 14:41:38 -0800 Subject: fs/binfmt_misc.c: add terminating newline to /proc/sys/fs/binfmt_misc/status The following is what it looks like before patching. It is not much readable. user@ubuntu:/proc/sys/fs/binfmt_misc$ cat status enableduser@ubuntu:/proc/sys/fs/binfmt_misc$ Signed-off-by: Qinghuang Feng Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/binfmt_misc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index e1158cb4fbd..c4e83537ead 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -649,7 +649,7 @@ static const struct file_operations bm_register_operations = { static ssize_t bm_status_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) { - char *s = enabled ? "enabled" : "disabled"; + char *s = enabled ? "enabled\n" : "disabled\n"; return simple_read_from_buffer(buf, nbytes, ppos, s, strlen(s)); } -- cgit v1.2.3-70-g09d2 From ee9ef6b778fbe1cacbec4fcd18a93f322ff93354 Mon Sep 17 00:00:00 2001 From: Qinghuang Feng Date: Tue, 6 Jan 2009 14:41:39 -0800 Subject: fs/ncpfs/getopt.c: cleanup keneldoc There are no argument named @flag in ncp_getopt(), remove it. Signed-off-by: Qinghuang Feng Cc: Randy Dunlap Cc: Petr Vandrovec Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ncpfs/getopt.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/ncpfs/getopt.c b/fs/ncpfs/getopt.c index 335b003dddf..0af3349de85 100644 --- a/fs/ncpfs/getopt.c +++ b/fs/ncpfs/getopt.c @@ -16,7 +16,6 @@ * @opts: an array of &struct option entries controlling parser operations * @optopt: output; will contain the current option * @optarg: output; will contain the value (if one exists) - * @flag: output; may be NULL; should point to a long for or'ing flags * @value: output; may be NULL; will be overwritten with the integer value * of the current argument. * -- cgit v1.2.3-70-g09d2 From 9c79f34f7ee71cd28272332b424ca64b2be006ab Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:41:57 -0800 Subject: eCryptfs: Filename Encryption: Tag 70 packets This patchset implements filename encryption via a passphrase-derived mount-wide Filename Encryption Key (FNEK) specified as a mount parameter. Each encrypted filename has a fixed prefix indicating that eCryptfs should try to decrypt the filename. When eCryptfs encounters this prefix, it decodes the filename into a tag 70 packet and then decrypts the packet contents using the FNEK, setting the filename to the decrypted filename. Both unencrypted and encrypted filenames can reside in the same lower filesystem. Because filename encryption expands the length of the filename during the encoding stage, eCryptfs will not properly handle filenames that are already near the maximum filename length. In the present implementation, eCryptfs must be able to produce a match against the lower encrypted and encoded filename representation when given a plaintext filename. Therefore, two files having the same plaintext name will encrypt and encode into the same lower filename if they are both encrypted using the same FNEK. This can be changed by finding a way to replace the prepended bytes in the blocked-aligned filename with random characters; they are hashes of the FNEK right now, so that it is possible to deterministically map from a plaintext filename to an encrypted and encoded filename in the lower filesystem. An implementation using random characters will have to decode and decrypt every single directory entry in any given directory any time an event occurs wherein the VFS needs to determine whether a particular file exists in the lower directory and the decrypted and decoded filenames have not yet been extracted for that directory. Thanks to Tyler Hicks and David Kleikamp for assistance in the development of this patchset. This patch: A tag 70 packet contains a filename encrypted with a Filename Encryption Key (FNEK). This patch implements functions for writing and parsing tag 70 packets. This patch also adds definitions and extends structures to support filename encryption. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 11 +- fs/ecryptfs/ecryptfs_kernel.h | 38 ++- fs/ecryptfs/keystore.c | 634 +++++++++++++++++++++++++++++++++++++----- 3 files changed, 613 insertions(+), 70 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 6046239465a..485732751f0 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1149,19 +1149,20 @@ ecryptfs_cipher_code_str_map[] = { /** * ecryptfs_code_for_cipher_string - * @crypt_stat: The cryptographic context + * @cipher_name: The string alias for the cipher + * @key_bytes: Length of key in bytes; used for AES code selection * * Returns zero on no match, or the cipher code on match */ -u8 ecryptfs_code_for_cipher_string(struct ecryptfs_crypt_stat *crypt_stat) +u8 ecryptfs_code_for_cipher_string(char *cipher_name, size_t key_bytes) { int i; u8 code = 0; struct ecryptfs_cipher_code_str_map_elem *map = ecryptfs_cipher_code_str_map; - if (strcmp(crypt_stat->cipher, "aes") == 0) { - switch (crypt_stat->key_size) { + if (strcmp(cipher_name, "aes") == 0) { + switch (key_bytes) { case 16: code = RFC2440_CIPHER_AES_128; break; @@ -1173,7 +1174,7 @@ u8 ecryptfs_code_for_cipher_string(struct ecryptfs_crypt_stat *crypt_stat) } } else { for (i = 0; i < ARRAY_SIZE(ecryptfs_cipher_code_str_map); i++) - if (strcmp(crypt_stat->cipher, map[i].cipher_str) == 0){ + if (strcmp(cipher_name, map[i].cipher_str) == 0) { code = map[i].cipher_code; break; } diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index a75026d35d1..76a95bd8819 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -199,6 +199,7 @@ ecryptfs_get_key_payload_data(struct key *key) #define ECRYPTFS_DEFAULT_CIPHER "aes" #define ECRYPTFS_DEFAULT_KEY_BYTES 16 #define ECRYPTFS_DEFAULT_HASH "md5" +#define ECRYPTFS_TAG_70_DIGEST ECRYPTFS_DEFAULT_HASH #define ECRYPTFS_TAG_1_PACKET_TYPE 0x01 #define ECRYPTFS_TAG_3_PACKET_TYPE 0x8C #define ECRYPTFS_TAG_11_PACKET_TYPE 0xED @@ -206,7 +207,25 @@ ecryptfs_get_key_payload_data(struct key *key) #define ECRYPTFS_TAG_65_PACKET_TYPE 0x41 #define ECRYPTFS_TAG_66_PACKET_TYPE 0x42 #define ECRYPTFS_TAG_67_PACKET_TYPE 0x43 +#define ECRYPTFS_TAG_70_PACKET_TYPE 0x46 /* FNEK-encrypted filename + * as dentry name */ +#define ECRYPTFS_TAG_71_PACKET_TYPE 0x47 /* FNEK-encrypted filename in + * metadata */ +#define ECRYPTFS_TAG_72_PACKET_TYPE 0x48 /* FEK-encrypted filename as + * dentry name */ +#define ECRYPTFS_TAG_73_PACKET_TYPE 0x49 /* FEK-encrypted filename as + * metadata */ +/* Constraint: ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES >= + * ECRYPTFS_MAX_IV_BYTES */ +#define ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES 16 +#define ECRYPTFS_NON_NULL 0x42 /* A reasonable substitute for NULL */ #define MD5_DIGEST_SIZE 16 +#define ECRYPTFS_TAG_70_DIGEST_SIZE MD5_DIGEST_SIZE +#define ECRYPTFS_FEK_ENCRYPTED_FILENAME_PREFIX "ECRYPTFS_FEK_ENCRYPTED." +#define ECRYPTFS_FEK_ENCRYPTED_FILENAME_PREFIX_SIZE 23 +#define ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX "ECRYPTFS_FNEK_ENCRYPTED." +#define ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE 24 +#define ECRYPTFS_ENCRYPTED_DENTRY_NAME_LEN (18 + 1 + 4 + 1 + 32) struct ecryptfs_key_sig { struct list_head crypt_stat_list; @@ -332,13 +351,20 @@ struct ecryptfs_mount_crypt_stat { #define ECRYPTFS_XATTR_METADATA_ENABLED 0x00000002 #define ECRYPTFS_ENCRYPTED_VIEW_ENABLED 0x00000004 #define ECRYPTFS_MOUNT_CRYPT_STAT_INITIALIZED 0x00000008 +#define ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES 0x00000010 +#define ECRYPTFS_GLOBAL_ENCFN_USE_MOUNT_FNEK 0x00000020 +#define ECRYPTFS_GLOBAL_ENCFN_USE_FEK 0x00000040 u32 flags; struct list_head global_auth_tok_list; struct mutex global_auth_tok_list_mutex; size_t num_global_auth_toks; size_t global_default_cipher_key_size; + size_t global_default_fn_cipher_key_bytes; unsigned char global_default_cipher_name[ECRYPTFS_MAX_CIPHER_NAME_SIZE + 1]; + unsigned char global_default_fn_cipher_name[ + ECRYPTFS_MAX_CIPHER_NAME_SIZE + 1]; + char global_default_fnek_sig[ECRYPTFS_SIG_SIZE_HEX + 1]; }; /* superblock private data. */ @@ -599,7 +625,7 @@ int ecryptfs_read_and_validate_header_region(char *data, struct inode *ecryptfs_inode); int ecryptfs_read_and_validate_xattr_region(char *page_virt, struct dentry *ecryptfs_dentry); -u8 ecryptfs_code_for_cipher_string(struct ecryptfs_crypt_stat *crypt_stat); +u8 ecryptfs_code_for_cipher_string(char *cipher_name, size_t key_bytes); int ecryptfs_cipher_code_to_string(char *str, u8 cipher_code); void ecryptfs_set_default_sizes(struct ecryptfs_crypt_stat *crypt_stat); int ecryptfs_generate_key_packet_set(char *dest_base, @@ -694,5 +720,15 @@ int ecryptfs_privileged_open(struct file **lower_file, struct vfsmount *lower_mnt, const struct cred *cred); int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry); +int +ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, + size_t *packet_size, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + char *filename, size_t filename_size); +int +ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, + size_t *packet_size, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + char *data, size_t max_packet_size); #endif /* #ifndef ECRYPTFS_KERNEL_H */ diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 0d713b69194..dafceb5560e 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -402,6 +402,569 @@ out: return rc; } +static int +ecryptfs_find_global_auth_tok_for_sig( + struct ecryptfs_global_auth_tok **global_auth_tok, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, char *sig) +{ + struct ecryptfs_global_auth_tok *walker; + int rc = 0; + + (*global_auth_tok) = NULL; + mutex_lock(&mount_crypt_stat->global_auth_tok_list_mutex); + list_for_each_entry(walker, + &mount_crypt_stat->global_auth_tok_list, + mount_crypt_stat_list) { + if (memcmp(walker->sig, sig, ECRYPTFS_SIG_SIZE_HEX) == 0) { + (*global_auth_tok) = walker; + goto out; + } + } + rc = -EINVAL; +out: + mutex_unlock(&mount_crypt_stat->global_auth_tok_list_mutex); + return rc; +} + +/** + * ecryptfs_find_auth_tok_for_sig + * @auth_tok: Set to the matching auth_tok; NULL if not found + * @crypt_stat: inode crypt_stat crypto context + * @sig: Sig of auth_tok to find + * + * For now, this function simply looks at the registered auth_tok's + * linked off the mount_crypt_stat, so all the auth_toks that can be + * used must be registered at mount time. This function could + * potentially try a lot harder to find auth_tok's (e.g., by calling + * out to ecryptfsd to dynamically retrieve an auth_tok object) so + * that static registration of auth_tok's will no longer be necessary. + * + * Returns zero on no error; non-zero on error + */ +static int +ecryptfs_find_auth_tok_for_sig( + struct ecryptfs_auth_tok **auth_tok, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + char *sig) +{ + struct ecryptfs_global_auth_tok *global_auth_tok; + int rc = 0; + + (*auth_tok) = NULL; + if (ecryptfs_find_global_auth_tok_for_sig(&global_auth_tok, + mount_crypt_stat, sig)) { + struct key *auth_tok_key; + + rc = ecryptfs_keyring_auth_tok_for_sig(&auth_tok_key, auth_tok, + sig); + } else + (*auth_tok) = global_auth_tok->global_auth_tok; + return rc; +} + +/** + * write_tag_70_packet can gobble a lot of stack space. We stuff most + * of the function's parameters in a kmalloc'd struct to help reduce + * eCryptfs' overall stack usage. + */ +struct ecryptfs_write_tag_70_packet_silly_stack { + u8 cipher_code; + size_t max_packet_size; + size_t packet_size_len; + size_t block_aligned_filename_size; + size_t block_size; + size_t i; + size_t j; + size_t num_rand_bytes; + struct mutex *tfm_mutex; + char *block_aligned_filename; + struct ecryptfs_auth_tok *auth_tok; + struct scatterlist src_sg; + struct scatterlist dst_sg; + struct blkcipher_desc desc; + char iv[ECRYPTFS_MAX_IV_BYTES]; + char hash[ECRYPTFS_TAG_70_DIGEST_SIZE]; + char tmp_hash[ECRYPTFS_TAG_70_DIGEST_SIZE]; + struct hash_desc hash_desc; + struct scatterlist hash_sg; +}; + +/** + * write_tag_70_packet - Write encrypted filename (EFN) packet against FNEK + * @filename: NULL-terminated filename string + * + * This is the simplest mechanism for achieving filename encryption in + * eCryptfs. It encrypts the given filename with the mount-wide + * filename encryption key (FNEK) and stores it in a packet to @dest, + * which the callee will encode and write directly into the dentry + * name. + */ +int +ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, + size_t *packet_size, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + char *filename, size_t filename_size) +{ + struct ecryptfs_write_tag_70_packet_silly_stack *s; + int rc = 0; + + s = kmalloc(sizeof(*s), GFP_KERNEL); + if (!s) { + printk(KERN_ERR "%s: Out of memory whilst trying to kmalloc " + "[%d] bytes of kernel memory\n", __func__, sizeof(*s)); + goto out; + } + s->desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; + (*packet_size) = 0; + rc = ecryptfs_get_tfm_and_mutex_for_cipher_name( + &s->desc.tfm, + &s->tfm_mutex, mount_crypt_stat->global_default_fn_cipher_name); + if (unlikely(rc)) { + printk(KERN_ERR "Internal error whilst attempting to get " + "tfm and mutex for cipher name [%s]; rc = [%d]\n", + mount_crypt_stat->global_default_fn_cipher_name, rc); + goto out; + } + mutex_lock(s->tfm_mutex); + s->block_size = crypto_blkcipher_blocksize(s->desc.tfm); + /* Plus one for the \0 separator between the random prefix + * and the plaintext filename */ + s->num_rand_bytes = (ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES + 1); + s->block_aligned_filename_size = (s->num_rand_bytes + filename_size); + if ((s->block_aligned_filename_size % s->block_size) != 0) { + s->num_rand_bytes += (s->block_size + - (s->block_aligned_filename_size + % s->block_size)); + s->block_aligned_filename_size = (s->num_rand_bytes + + filename_size); + } + /* Octet 0: Tag 70 identifier + * Octets 1-N1: Tag 70 packet size (includes cipher identifier + * and block-aligned encrypted filename size) + * Octets N1-N2: FNEK sig (ECRYPTFS_SIG_SIZE) + * Octet N2-N3: Cipher identifier (1 octet) + * Octets N3-N4: Block-aligned encrypted filename + * - Consists of a minimum number of random characters, a \0 + * separator, and then the filename */ + s->max_packet_size = (1 /* Tag 70 identifier */ + + 3 /* Max Tag 70 packet size */ + + ECRYPTFS_SIG_SIZE /* FNEK sig */ + + 1 /* Cipher identifier */ + + s->block_aligned_filename_size); + if (dest == NULL) { + (*packet_size) = s->max_packet_size; + goto out_unlock; + } + if (s->max_packet_size > (*remaining_bytes)) { + printk(KERN_WARNING "%s: Require [%d] bytes to write; only " + "[%d] available\n", __func__, s->max_packet_size, + (*remaining_bytes)); + rc = -EINVAL; + goto out_unlock; + } + s->block_aligned_filename = kzalloc(s->block_aligned_filename_size, + GFP_KERNEL); + if (!s->block_aligned_filename) { + printk(KERN_ERR "%s: Out of kernel memory whilst attempting to " + "kzalloc [%Zd] bytes\n", __func__, + s->block_aligned_filename_size); + rc = -ENOMEM; + goto out_unlock; + } + s->i = 0; + dest[s->i++] = ECRYPTFS_TAG_70_PACKET_TYPE; + rc = ecryptfs_write_packet_length(&dest[s->i], + (ECRYPTFS_SIG_SIZE + + 1 /* Cipher code */ + + s->block_aligned_filename_size), + &s->packet_size_len); + if (rc) { + printk(KERN_ERR "%s: Error generating tag 70 packet " + "header; cannot generate packet length; rc = [%d]\n", + __func__, rc); + goto out_free_unlock; + } + s->i += s->packet_size_len; + ecryptfs_from_hex(&dest[s->i], + mount_crypt_stat->global_default_fnek_sig, + ECRYPTFS_SIG_SIZE); + s->i += ECRYPTFS_SIG_SIZE; + s->cipher_code = ecryptfs_code_for_cipher_string( + mount_crypt_stat->global_default_fn_cipher_name, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + if (s->cipher_code == 0) { + printk(KERN_WARNING "%s: Unable to generate code for " + "cipher [%s] with key bytes [%d]\n", __func__, + mount_crypt_stat->global_default_fn_cipher_name, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + rc = -EINVAL; + goto out_free_unlock; + } + dest[s->i++] = s->cipher_code; + rc = ecryptfs_find_auth_tok_for_sig( + &s->auth_tok, mount_crypt_stat, + mount_crypt_stat->global_default_fnek_sig); + if (rc) { + printk(KERN_ERR "%s: Error attempting to find auth tok for " + "fnek sig [%s]; rc = [%d]\n", __func__, + mount_crypt_stat->global_default_fnek_sig, rc); + goto out_free_unlock; + } + /* TODO: Support other key modules than passphrase for + * filename encryption */ + BUG_ON(s->auth_tok->token_type != ECRYPTFS_PASSWORD); + sg_init_one( + &s->hash_sg, + (u8 *)s->auth_tok->token.password.session_key_encryption_key, + s->auth_tok->token.password.session_key_encryption_key_bytes); + s->hash_desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; + s->hash_desc.tfm = crypto_alloc_hash(ECRYPTFS_TAG_70_DIGEST, 0, + CRYPTO_ALG_ASYNC); + if (IS_ERR(s->hash_desc.tfm)) { + rc = PTR_ERR(s->hash_desc.tfm); + printk(KERN_ERR "%s: Error attempting to " + "allocate hash crypto context; rc = [%d]\n", + __func__, rc); + goto out_free_unlock; + } + rc = crypto_hash_init(&s->hash_desc); + if (rc) { + printk(KERN_ERR + "%s: Error initializing crypto hash; rc = [%d]\n", + __func__, rc); + goto out_release_free_unlock; + } + rc = crypto_hash_update( + &s->hash_desc, &s->hash_sg, + s->auth_tok->token.password.session_key_encryption_key_bytes); + if (rc) { + printk(KERN_ERR + "%s: Error updating crypto hash; rc = [%d]\n", + __func__, rc); + goto out_release_free_unlock; + } + rc = crypto_hash_final(&s->hash_desc, s->hash); + if (rc) { + printk(KERN_ERR + "%s: Error finalizing crypto hash; rc = [%d]\n", + __func__, rc); + goto out_release_free_unlock; + } + for (s->j = 0; s->j < (s->num_rand_bytes - 1); s->j++) { + s->block_aligned_filename[s->j] = + s->hash[(s->j % ECRYPTFS_TAG_70_DIGEST_SIZE)]; + if ((s->j % ECRYPTFS_TAG_70_DIGEST_SIZE) + == (ECRYPTFS_TAG_70_DIGEST_SIZE - 1)) { + sg_init_one(&s->hash_sg, (u8 *)s->hash, + ECRYPTFS_TAG_70_DIGEST_SIZE); + rc = crypto_hash_init(&s->hash_desc); + if (rc) { + printk(KERN_ERR + "%s: Error initializing crypto hash; " + "rc = [%d]\n", __func__, rc); + goto out_release_free_unlock; + } + rc = crypto_hash_update(&s->hash_desc, &s->hash_sg, + ECRYPTFS_TAG_70_DIGEST_SIZE); + if (rc) { + printk(KERN_ERR + "%s: Error updating crypto hash; " + "rc = [%d]\n", __func__, rc); + goto out_release_free_unlock; + } + rc = crypto_hash_final(&s->hash_desc, s->tmp_hash); + if (rc) { + printk(KERN_ERR + "%s: Error finalizing crypto hash; " + "rc = [%d]\n", __func__, rc); + goto out_release_free_unlock; + } + memcpy(s->hash, s->tmp_hash, + ECRYPTFS_TAG_70_DIGEST_SIZE); + } + if (s->block_aligned_filename[s->j] == '\0') + s->block_aligned_filename[s->j] = ECRYPTFS_NON_NULL; + } + memcpy(&s->block_aligned_filename[s->num_rand_bytes], filename, + filename_size); + rc = virt_to_scatterlist(s->block_aligned_filename, + s->block_aligned_filename_size, &s->src_sg, 1); + if (rc != 1) { + printk(KERN_ERR "%s: Internal error whilst attempting to " + "convert filename memory to scatterlist; " + "expected rc = 1; got rc = [%d]. " + "block_aligned_filename_size = [%d]\n", __func__, rc, + s->block_aligned_filename_size); + goto out_release_free_unlock; + } + rc = virt_to_scatterlist(&dest[s->i], s->block_aligned_filename_size, + &s->dst_sg, 1); + if (rc != 1) { + printk(KERN_ERR "%s: Internal error whilst attempting to " + "convert encrypted filename memory to scatterlist; " + "expected rc = 1; got rc = [%d]. " + "block_aligned_filename_size = [%d]\n", __func__, rc, + s->block_aligned_filename_size); + goto out_release_free_unlock; + } + /* The characters in the first block effectively do the job + * of the IV here, so we just use 0's for the IV. Note the + * constraint that ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES + * >= ECRYPTFS_MAX_IV_BYTES. */ + memset(s->iv, 0, ECRYPTFS_MAX_IV_BYTES); + s->desc.info = s->iv; + rc = crypto_blkcipher_setkey( + s->desc.tfm, + s->auth_tok->token.password.session_key_encryption_key, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + if (rc < 0) { + printk(KERN_ERR "%s: Error setting key for crypto context; " + "rc = [%d]. s->auth_tok->token.password.session_key_" + "encryption_key = [0x%p]; mount_crypt_stat->" + "global_default_fn_cipher_key_bytes = [%Zd]\n", __func__, + rc, + s->auth_tok->token.password.session_key_encryption_key, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + goto out_release_free_unlock; + } + rc = crypto_blkcipher_encrypt_iv(&s->desc, &s->dst_sg, &s->src_sg, + s->block_aligned_filename_size); + if (rc) { + printk(KERN_ERR "%s: Error attempting to encrypt filename; " + "rc = [%d]\n", __func__, rc); + goto out_release_free_unlock; + } + s->i += s->block_aligned_filename_size; + (*packet_size) = s->i; + (*remaining_bytes) -= (*packet_size); +out_release_free_unlock: + crypto_free_hash(s->hash_desc.tfm); +out_free_unlock: + memset(s->block_aligned_filename, 0, s->block_aligned_filename_size); + kfree(s->block_aligned_filename); +out_unlock: + mutex_unlock(s->tfm_mutex); +out: + kfree(s); + return rc; +} + +struct ecryptfs_parse_tag_70_packet_silly_stack { + u8 cipher_code; + size_t max_packet_size; + size_t packet_size_len; + size_t parsed_tag_70_packet_size; + size_t block_aligned_filename_size; + size_t block_size; + size_t i; + struct mutex *tfm_mutex; + char *decrypted_filename; + struct ecryptfs_auth_tok *auth_tok; + struct scatterlist src_sg; + struct scatterlist dst_sg; + struct blkcipher_desc desc; + char fnek_sig_hex[ECRYPTFS_SIG_SIZE_HEX + 1]; + char iv[ECRYPTFS_MAX_IV_BYTES]; + char cipher_string[ECRYPTFS_MAX_CIPHER_NAME_SIZE]; +}; + +/** + * parse_tag_70_packet - Parse and process FNEK-encrypted passphrase packet + * @filename: This function kmalloc's the memory for the filename + */ +int +ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, + size_t *packet_size, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + char *data, size_t max_packet_size) +{ + struct ecryptfs_parse_tag_70_packet_silly_stack *s; + int rc = 0; + + (*packet_size) = 0; + (*filename_size) = 0; + (*filename) = NULL; + s = kmalloc(sizeof(*s), GFP_KERNEL); + if (!s) { + printk(KERN_ERR "%s: Out of memory whilst trying to kmalloc " + "[%d] bytes of kernel memory\n", __func__, sizeof(*s)); + goto out; + } + s->desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; + if (max_packet_size < (1 + 1 + ECRYPTFS_SIG_SIZE + 1 + 1)) { + printk(KERN_WARNING "%s: max_packet_size is [%Zd]; it must be " + "at least [%d]\n", __func__, max_packet_size, + (1 + 1 + ECRYPTFS_SIG_SIZE + 1 + 1)); + rc = -EINVAL; + goto out; + } + /* Octet 0: Tag 70 identifier + * Octets 1-N1: Tag 70 packet size (includes cipher identifier + * and block-aligned encrypted filename size) + * Octets N1-N2: FNEK sig (ECRYPTFS_SIG_SIZE) + * Octet N2-N3: Cipher identifier (1 octet) + * Octets N3-N4: Block-aligned encrypted filename + * - Consists of a minimum number of random numbers, a \0 + * separator, and then the filename */ + if (data[(*packet_size)++] != ECRYPTFS_TAG_70_PACKET_TYPE) { + printk(KERN_WARNING "%s: Invalid packet tag [0x%.2x]; must be " + "tag [0x%.2x]\n", __func__, + data[((*packet_size) - 1)], ECRYPTFS_TAG_70_PACKET_TYPE); + rc = -EINVAL; + goto out; + } + rc = ecryptfs_parse_packet_length(&data[(*packet_size)], + &s->parsed_tag_70_packet_size, + &s->packet_size_len); + if (rc) { + printk(KERN_WARNING "%s: Error parsing packet length; " + "rc = [%d]\n", __func__, rc); + goto out; + } + s->block_aligned_filename_size = (s->parsed_tag_70_packet_size + - ECRYPTFS_SIG_SIZE - 1); + if ((1 + s->packet_size_len + s->parsed_tag_70_packet_size) + > max_packet_size) { + printk(KERN_WARNING "%s: max_packet_size is [%d]; real packet " + "size is [%d]\n", __func__, max_packet_size, + (1 + s->packet_size_len + 1 + + s->block_aligned_filename_size)); + rc = -EINVAL; + goto out; + } + (*packet_size) += s->packet_size_len; + ecryptfs_to_hex(s->fnek_sig_hex, &data[(*packet_size)], + ECRYPTFS_SIG_SIZE); + s->fnek_sig_hex[ECRYPTFS_SIG_SIZE_HEX] = '\0'; + (*packet_size) += ECRYPTFS_SIG_SIZE; + s->cipher_code = data[(*packet_size)++]; + rc = ecryptfs_cipher_code_to_string(s->cipher_string, s->cipher_code); + if (rc) { + printk(KERN_WARNING "%s: Cipher code [%d] is invalid\n", + __func__, s->cipher_code); + goto out; + } + rc = ecryptfs_get_tfm_and_mutex_for_cipher_name(&s->desc.tfm, + &s->tfm_mutex, + s->cipher_string); + if (unlikely(rc)) { + printk(KERN_ERR "Internal error whilst attempting to get " + "tfm and mutex for cipher name [%s]; rc = [%d]\n", + s->cipher_string, rc); + goto out; + } + mutex_lock(s->tfm_mutex); + rc = virt_to_scatterlist(&data[(*packet_size)], + s->block_aligned_filename_size, &s->src_sg, 1); + if (rc != 1) { + printk(KERN_ERR "%s: Internal error whilst attempting to " + "convert encrypted filename memory to scatterlist; " + "expected rc = 1; got rc = [%d]. " + "block_aligned_filename_size = [%d]\n", __func__, rc, + s->block_aligned_filename_size); + goto out_unlock; + } + (*packet_size) += s->block_aligned_filename_size; + s->decrypted_filename = kmalloc(s->block_aligned_filename_size, + GFP_KERNEL); + if (!s->decrypted_filename) { + printk(KERN_ERR "%s: Out of memory whilst attempting to " + "kmalloc [%d] bytes\n", __func__, + s->block_aligned_filename_size); + rc = -ENOMEM; + goto out_unlock; + } + rc = virt_to_scatterlist(s->decrypted_filename, + s->block_aligned_filename_size, &s->dst_sg, 1); + if (rc != 1) { + printk(KERN_ERR "%s: Internal error whilst attempting to " + "convert decrypted filename memory to scatterlist; " + "expected rc = 1; got rc = [%d]. " + "block_aligned_filename_size = [%d]\n", __func__, rc, + s->block_aligned_filename_size); + goto out_free_unlock; + } + /* The characters in the first block effectively do the job of + * the IV here, so we just use 0's for the IV. Note the + * constraint that ECRYPTFS_FILENAME_MIN_RANDOM_PREPEND_BYTES + * >= ECRYPTFS_MAX_IV_BYTES. */ + memset(s->iv, 0, ECRYPTFS_MAX_IV_BYTES); + s->desc.info = s->iv; + rc = ecryptfs_find_auth_tok_for_sig(&s->auth_tok, mount_crypt_stat, + s->fnek_sig_hex); + if (rc) { + printk(KERN_ERR "%s: Error attempting to find auth tok for " + "fnek sig [%s]; rc = [%d]\n", __func__, s->fnek_sig_hex, + rc); + goto out_free_unlock; + } + /* TODO: Support other key modules than passphrase for + * filename encryption */ + BUG_ON(s->auth_tok->token_type != ECRYPTFS_PASSWORD); + rc = crypto_blkcipher_setkey( + s->desc.tfm, + s->auth_tok->token.password.session_key_encryption_key, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + if (rc < 0) { + printk(KERN_ERR "%s: Error setting key for crypto context; " + "rc = [%d]. s->auth_tok->token.password.session_key_" + "encryption_key = [0x%p]; mount_crypt_stat->" + "global_default_fn_cipher_key_bytes = [%Zd]\n", __func__, + rc, + s->auth_tok->token.password.session_key_encryption_key, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + goto out_free_unlock; + } + rc = crypto_blkcipher_decrypt_iv(&s->desc, &s->dst_sg, &s->src_sg, + s->block_aligned_filename_size); + if (rc) { + printk(KERN_ERR "%s: Error attempting to decrypt filename; " + "rc = [%d]\n", __func__, rc); + goto out_free_unlock; + } + s->i = 0; + while (s->decrypted_filename[s->i] != '\0' + && s->i < s->block_aligned_filename_size) + s->i++; + if (s->i == s->block_aligned_filename_size) { + printk(KERN_WARNING "%s: Invalid tag 70 packet; could not " + "find valid separator between random characters and " + "the filename\n", __func__); + rc = -EINVAL; + goto out_free_unlock; + } + s->i++; + (*filename_size) = (s->block_aligned_filename_size - s->i); + if (!((*filename_size) > 0 && (*filename_size < PATH_MAX))) { + printk(KERN_WARNING "%s: Filename size is [%Zd], which is " + "invalid\n", __func__, (*filename_size)); + rc = -EINVAL; + goto out_free_unlock; + } + (*filename) = kmalloc(((*filename_size) + 1), GFP_KERNEL); + if (!(*filename)) { + printk(KERN_ERR "%s: Out of memory whilst attempting to " + "kmalloc [%d] bytes\n", __func__, + ((*filename_size) + 1)); + rc = -ENOMEM; + goto out_free_unlock; + } + memcpy((*filename), &s->decrypted_filename[s->i], (*filename_size)); + (*filename)[(*filename_size)] = '\0'; +out_free_unlock: + kfree(s->decrypted_filename); +out_unlock: + mutex_unlock(s->tfm_mutex); +out: + if (rc) { + (*packet_size) = 0; + (*filename_size) = 0; + (*filename) = NULL; + } + kfree(s); + return rc; +} + static int ecryptfs_get_auth_tok_sig(char **sig, struct ecryptfs_auth_tok *auth_tok) { @@ -897,30 +1460,6 @@ out: return rc; } -static int -ecryptfs_find_global_auth_tok_for_sig( - struct ecryptfs_global_auth_tok **global_auth_tok, - struct ecryptfs_mount_crypt_stat *mount_crypt_stat, char *sig) -{ - struct ecryptfs_global_auth_tok *walker; - int rc = 0; - - (*global_auth_tok) = NULL; - mutex_lock(&mount_crypt_stat->global_auth_tok_list_mutex); - list_for_each_entry(walker, - &mount_crypt_stat->global_auth_tok_list, - mount_crypt_stat_list) { - if (memcmp(walker->sig, sig, ECRYPTFS_SIG_SIZE_HEX) == 0) { - (*global_auth_tok) = walker; - goto out; - } - } - rc = -EINVAL; -out: - mutex_unlock(&mount_crypt_stat->global_auth_tok_list_mutex); - return rc; -} - /** * ecryptfs_verify_version * @version: The version number to confirm @@ -989,43 +1528,6 @@ out: return rc; } -/** - * ecryptfs_find_auth_tok_for_sig - * @auth_tok: Set to the matching auth_tok; NULL if not found - * @crypt_stat: inode crypt_stat crypto context - * @sig: Sig of auth_tok to find - * - * For now, this function simply looks at the registered auth_tok's - * linked off the mount_crypt_stat, so all the auth_toks that can be - * used must be registered at mount time. This function could - * potentially try a lot harder to find auth_tok's (e.g., by calling - * out to ecryptfsd to dynamically retrieve an auth_tok object) so - * that static registration of auth_tok's will no longer be necessary. - * - * Returns zero on no error; non-zero on error - */ -static int -ecryptfs_find_auth_tok_for_sig( - struct ecryptfs_auth_tok **auth_tok, - struct ecryptfs_crypt_stat *crypt_stat, char *sig) -{ - struct ecryptfs_mount_crypt_stat *mount_crypt_stat = - crypt_stat->mount_crypt_stat; - struct ecryptfs_global_auth_tok *global_auth_tok; - int rc = 0; - - (*auth_tok) = NULL; - if (ecryptfs_find_global_auth_tok_for_sig(&global_auth_tok, - mount_crypt_stat, sig)) { - struct key *auth_tok_key; - - rc = ecryptfs_keyring_auth_tok_for_sig(&auth_tok_key, auth_tok, - sig); - } else - (*auth_tok) = global_auth_tok->global_auth_tok; - return rc; -} - /** * decrypt_passphrase_encrypted_session_key - Decrypt the session key with the given auth_tok. * @auth_tok: The passphrase authentication token to use to encrypt the FEK @@ -1256,7 +1758,8 @@ find_next_matching_auth_tok: rc = -EINVAL; goto out_wipe_list; } - ecryptfs_find_auth_tok_for_sig(&matching_auth_tok, crypt_stat, + ecryptfs_find_auth_tok_for_sig(&matching_auth_tok, + crypt_stat->mount_crypt_stat, candidate_auth_tok_sig); if (matching_auth_tok) { found_auth_tok = 1; @@ -1336,7 +1839,9 @@ pki_encrypt_session_key(struct ecryptfs_auth_tok *auth_tok, int rc; rc = write_tag_66_packet(auth_tok->token.private_key.signature, - ecryptfs_code_for_cipher_string(crypt_stat), + ecryptfs_code_for_cipher_string( + crypt_stat->cipher, + crypt_stat->key_size), crypt_stat, &payload, &payload_len); if (rc) { ecryptfs_printk(KERN_ERR, "Error generating tag 66 packet\n"); @@ -1696,7 +2201,8 @@ encrypted_session_key_set: dest[(*packet_size)++] = 0x04; /* version 4 */ /* TODO: Break from RFC2440 so that arbitrary ciphers can be * specified with strings */ - cipher_code = ecryptfs_code_for_cipher_string(crypt_stat); + cipher_code = ecryptfs_code_for_cipher_string(crypt_stat->cipher, + crypt_stat->key_size); if (cipher_code == 0) { ecryptfs_printk(KERN_WARNING, "Unable to generate code for " "cipher [%s]\n", crypt_stat->cipher); -- cgit v1.2.3-70-g09d2 From a34f60f748c6fe5d791e9b54cffe442201428254 Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:41:58 -0800 Subject: eCryptfs: Filename Encryption: Header updates Extensions to the header file to support filename encryption. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 4 +-- fs/ecryptfs/ecryptfs_kernel.h | 61 ++++++++++++++++++++++++++++++++++--------- 2 files changed, 51 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 485732751f0..c9839df37c7 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -175,8 +175,8 @@ out: * * Returns zero on success; non-zero on error. */ -static int ecryptfs_derive_iv(char *iv, struct ecryptfs_crypt_stat *crypt_stat, - loff_t offset) +int ecryptfs_derive_iv(char *iv, struct ecryptfs_crypt_stat *crypt_stat, + loff_t offset) { int rc = 0; char dst[MD5_DIGEST_SIZE]; diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index 76a95bd8819..b648175a44c 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -51,12 +51,16 @@ #define ECRYPTFS_VERSIONING_XATTR 0x00000010 #define ECRYPTFS_VERSIONING_MULTKEY 0x00000020 #define ECRYPTFS_VERSIONING_DEVMISC 0x00000040 +#define ECRYPTFS_VERSIONING_HMAC 0x00000080 +#define ECRYPTFS_VERSIONING_FILENAME_ENCRYPTION 0x00000100 +#define ECRYPTFS_VERSIONING_GCM 0x00000200 #define ECRYPTFS_VERSIONING_MASK (ECRYPTFS_VERSIONING_PASSPHRASE \ | ECRYPTFS_VERSIONING_PLAINTEXT_PASSTHROUGH \ | ECRYPTFS_VERSIONING_PUBKEY \ | ECRYPTFS_VERSIONING_XATTR \ | ECRYPTFS_VERSIONING_MULTKEY \ - | ECRYPTFS_VERSIONING_DEVMISC) + | ECRYPTFS_VERSIONING_DEVMISC \ + | ECRYPTFS_VERSIONING_FILENAME_ENCRYPTION) #define ECRYPTFS_MAX_PASSWORD_LENGTH 64 #define ECRYPTFS_MAX_PASSPHRASE_BYTES ECRYPTFS_MAX_PASSWORD_LENGTH #define ECRYPTFS_SALT_SIZE 8 @@ -232,23 +236,39 @@ struct ecryptfs_key_sig { char keysig[ECRYPTFS_SIG_SIZE_HEX]; }; +struct ecryptfs_filename { + struct list_head crypt_stat_list; +#define ECRYPTFS_FILENAME_CONTAINS_DECRYPTED 0x00000001 + u32 flags; + u32 seq_no; + char *filename; + char *encrypted_filename; + size_t filename_size; + size_t encrypted_filename_size; + char fnek_sig[ECRYPTFS_SIG_SIZE_HEX]; + char dentry_name[ECRYPTFS_ENCRYPTED_DENTRY_NAME_LEN + 1]; +}; + /** * This is the primary struct associated with each encrypted file. * * TODO: cache align/pack? */ struct ecryptfs_crypt_stat { -#define ECRYPTFS_STRUCT_INITIALIZED 0x00000001 -#define ECRYPTFS_POLICY_APPLIED 0x00000002 -#define ECRYPTFS_NEW_FILE 0x00000004 -#define ECRYPTFS_ENCRYPTED 0x00000008 -#define ECRYPTFS_SECURITY_WARNING 0x00000010 -#define ECRYPTFS_ENABLE_HMAC 0x00000020 -#define ECRYPTFS_ENCRYPT_IV_PAGES 0x00000040 -#define ECRYPTFS_KEY_VALID 0x00000080 -#define ECRYPTFS_METADATA_IN_XATTR 0x00000100 -#define ECRYPTFS_VIEW_AS_ENCRYPTED 0x00000200 -#define ECRYPTFS_KEY_SET 0x00000400 +#define ECRYPTFS_STRUCT_INITIALIZED 0x00000001 +#define ECRYPTFS_POLICY_APPLIED 0x00000002 +#define ECRYPTFS_NEW_FILE 0x00000004 +#define ECRYPTFS_ENCRYPTED 0x00000008 +#define ECRYPTFS_SECURITY_WARNING 0x00000010 +#define ECRYPTFS_ENABLE_HMAC 0x00000020 +#define ECRYPTFS_ENCRYPT_IV_PAGES 0x00000040 +#define ECRYPTFS_KEY_VALID 0x00000080 +#define ECRYPTFS_METADATA_IN_XATTR 0x00000100 +#define ECRYPTFS_VIEW_AS_ENCRYPTED 0x00000200 +#define ECRYPTFS_KEY_SET 0x00000400 +#define ECRYPTFS_ENCRYPT_FILENAMES 0x00000800 +#define ECRYPTFS_ENCFN_USE_MOUNT_FNEK 0x00001000 +#define ECRYPTFS_ENCFN_USE_FEK 0x00002000 u32 flags; unsigned int file_version; size_t iv_bytes; @@ -597,6 +617,15 @@ struct ecryptfs_open_req { int ecryptfs_interpose(struct dentry *hidden_dentry, struct dentry *this_dentry, struct super_block *sb, u32 flags); +int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, + struct dentry *lower_dentry, + struct ecryptfs_crypt_stat *crypt_stat, + struct inode *ecryptfs_dir_inode, + struct nameidata *ecryptfs_nd); +int ecryptfs_decode_and_decrypt_filename(char **decrypted_name, + size_t *decrypted_name_size, + struct dentry *ecryptfs_dentry, + const char *name, size_t name_size); int ecryptfs_fill_zeros(struct file *file, loff_t new_length); int ecryptfs_decode_filename(struct ecryptfs_crypt_stat *crypt_stat, const char *name, int length, @@ -604,6 +633,12 @@ int ecryptfs_decode_filename(struct ecryptfs_crypt_stat *crypt_stat, int ecryptfs_encode_filename(struct ecryptfs_crypt_stat *crypt_stat, const char *name, int length, char **encoded_name); +int ecryptfs_encrypt_and_encode_filename( + char **encoded_name, + size_t *encoded_name_size, + struct ecryptfs_crypt_stat *crypt_stat, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + const char *name, size_t name_size); struct dentry *ecryptfs_lower_dentry(struct dentry *this_dentry); void ecryptfs_dump_hex(char *data, int bytes); int virt_to_scatterlist(const void *addr, int size, struct scatterlist *sg, @@ -730,5 +765,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, size_t *packet_size, struct ecryptfs_mount_crypt_stat *mount_crypt_stat, char *data, size_t max_packet_size); +int ecryptfs_derive_iv(char *iv, struct ecryptfs_crypt_stat *crypt_stat, + loff_t offset); #endif /* #ifndef ECRYPTFS_KERNEL_H */ -- cgit v1.2.3-70-g09d2 From 51ca58dcc9f0d6b1e78954d08bd4954fb6a1421c Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:41:59 -0800 Subject: eCryptfs: Filename Encryption: Encoding and encryption functions These functions support encrypting and encoding the filename contents. The encrypted filename contents may consist of any ASCII characters. This patch includes a custom encoding mechanism to map the ASCII characters to a reduced character set that is appropriate for filenames. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 433 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 433 insertions(+) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index c9839df37c7..18c78abba68 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1720,6 +1720,98 @@ out: return error; } +/** + * ecryptfs_encrypt_filename - encrypt filename + * + * CBC-encrypts the filename. We do not want to encrypt the same + * filename with the same key and IV, which may happen with hard + * links, so we prepend random bits to each filename. + * + * Returns zero on success; non-zero otherwise + */ +static int +ecryptfs_encrypt_filename(struct ecryptfs_filename *filename, + struct ecryptfs_crypt_stat *crypt_stat, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat) +{ + int rc = 0; + + filename->encrypted_filename = NULL; + filename->encrypted_filename_size = 0; + if ((crypt_stat && (crypt_stat->flags & ECRYPTFS_ENCFN_USE_MOUNT_FNEK)) + || (mount_crypt_stat && (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCFN_USE_MOUNT_FNEK))) { + size_t packet_size; + size_t remaining_bytes; + + rc = ecryptfs_write_tag_70_packet( + NULL, NULL, + &filename->encrypted_filename_size, + mount_crypt_stat, NULL, + filename->filename_size); + if (rc) { + printk(KERN_ERR "%s: Error attempting to get packet " + "size for tag 72; rc = [%d]\n", __func__, + rc); + filename->encrypted_filename_size = 0; + goto out; + } + filename->encrypted_filename = + kmalloc(filename->encrypted_filename_size, GFP_KERNEL); + if (!filename->encrypted_filename) { + printk(KERN_ERR "%s: Out of memory whilst attempting " + "to kmalloc [%Zd] bytes\n", __func__, + filename->encrypted_filename_size); + rc = -ENOMEM; + goto out; + } + remaining_bytes = filename->encrypted_filename_size; + rc = ecryptfs_write_tag_70_packet(filename->encrypted_filename, + &remaining_bytes, + &packet_size, + mount_crypt_stat, + filename->filename, + filename->filename_size); + if (rc) { + printk(KERN_ERR "%s: Error attempting to generate " + "tag 70 packet; rc = [%d]\n", __func__, + rc); + kfree(filename->encrypted_filename); + filename->encrypted_filename = NULL; + filename->encrypted_filename_size = 0; + goto out; + } + filename->encrypted_filename_size = packet_size; + } else { + printk(KERN_ERR "%s: No support for requested filename " + "encryption method in this release\n", __func__); + rc = -ENOTSUPP; + goto out; + } +out: + return rc; +} + +static int ecryptfs_copy_filename(char **copied_name, size_t *copied_name_size, + const char *name, size_t name_size) +{ + int rc = 0; + + (*copied_name) = kmalloc((name_size + 2), GFP_KERNEL); + if (!(*copied_name)) { + rc = -ENOMEM; + goto out; + } + memcpy((void *)(*copied_name), (void *)name, name_size); + (*copied_name)[(name_size)] = '\0'; /* Only for convenience + * in printing out the + * string in debug + * messages */ + (*copied_name_size) = (name_size + 1); +out: + return rc; +} + /** * ecryptfs_process_key_cipher - Perform key cipher initialization. * @key_tfm: Crypto context for key material, set by this function @@ -1911,3 +2003,344 @@ out: mutex_unlock(&key_tfm_list_mutex); return rc; } + +/* 64 characters forming a 6-bit target field */ +static unsigned char *portable_filename_chars = ("-.0123456789ABCD" + "EFGHIJKLMNOPQRST" + "UVWXYZabcdefghij" + "klmnopqrstuvwxyz"); + +/* We could either offset on every reverse map or just pad some 0x00's + * at the front here */ +static unsigned char filename_rev_map[] = { + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 7 */ + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 15 */ + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 23 */ + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 31 */ + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 39 */ + 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, /* 47 */ + 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, /* 55 */ + 0x0A, 0x0B, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 63 */ + 0x00, 0x0C, 0x0D, 0x0E, 0x0F, 0x10, 0x11, 0x12, /* 71 */ + 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19, 0x1A, /* 79 */ + 0x1B, 0x1C, 0x1D, 0x1E, 0x1F, 0x20, 0x21, 0x22, /* 87 */ + 0x23, 0x24, 0x25, 0x00, 0x00, 0x00, 0x00, 0x00, /* 95 */ + 0x00, 0x26, 0x27, 0x28, 0x29, 0x2A, 0x2B, 0x2C, /* 103 */ + 0x2D, 0x2E, 0x2F, 0x30, 0x31, 0x32, 0x33, 0x34, /* 111 */ + 0x35, 0x36, 0x37, 0x38, 0x39, 0x3A, 0x3B, 0x3C, /* 119 */ + 0x3D, 0x3E, 0x3F +}; + +/** + * ecryptfs_encode_for_filename + * @dst: Destination location for encoded filename + * @dst_size: Size of the encoded filename in bytes + * @src: Source location for the filename to encode + * @src_size: Size of the source in bytes + */ +void ecryptfs_encode_for_filename(unsigned char *dst, size_t *dst_size, + unsigned char *src, size_t src_size) +{ + size_t num_blocks; + size_t block_num = 0; + size_t dst_offset = 0; + unsigned char last_block[3]; + + if (src_size == 0) { + (*dst_size) = 0; + goto out; + } + num_blocks = (src_size / 3); + if ((src_size % 3) == 0) { + memcpy(last_block, (&src[src_size - 3]), 3); + } else { + num_blocks++; + last_block[2] = 0x00; + switch (src_size % 3) { + case 1: + last_block[0] = src[src_size - 1]; + last_block[1] = 0x00; + break; + case 2: + last_block[0] = src[src_size - 2]; + last_block[1] = src[src_size - 1]; + } + } + (*dst_size) = (num_blocks * 4); + if (!dst) + goto out; + while (block_num < num_blocks) { + unsigned char *src_block; + unsigned char dst_block[4]; + + if (block_num == (num_blocks - 1)) + src_block = last_block; + else + src_block = &src[block_num * 3]; + dst_block[0] = ((src_block[0] >> 2) & 0x3F); + dst_block[1] = (((src_block[0] << 4) & 0x30) + | ((src_block[1] >> 4) & 0x0F)); + dst_block[2] = (((src_block[1] << 2) & 0x3C) + | ((src_block[2] >> 6) & 0x03)); + dst_block[3] = (src_block[2] & 0x3F); + dst[dst_offset++] = portable_filename_chars[dst_block[0]]; + dst[dst_offset++] = portable_filename_chars[dst_block[1]]; + dst[dst_offset++] = portable_filename_chars[dst_block[2]]; + dst[dst_offset++] = portable_filename_chars[dst_block[3]]; + block_num++; + } +out: + return; +} + +int ecryptfs_decode_from_filename(unsigned char *dst, size_t *dst_size, + const unsigned char *src, size_t src_size) +{ + u8 current_bit_offset = 0; + size_t src_byte_offset = 0; + size_t dst_byte_offset = 0; + int rc = 0; + + if (dst == NULL) { + /* Not exact; conservatively long */ + (*dst_size) = (((src_size + 1) * 3) / 4); + goto out; + } + while (src_byte_offset < src_size) { + unsigned char src_byte = + filename_rev_map[(int)src[src_byte_offset]]; + + switch (current_bit_offset) { + case 0: + dst[dst_byte_offset] = (src_byte << 2); + current_bit_offset = 6; + break; + case 6: + dst[dst_byte_offset++] |= (src_byte >> 4); + dst[dst_byte_offset] = ((src_byte & 0xF) + << 4); + current_bit_offset = 4; + break; + case 4: + dst[dst_byte_offset++] |= (src_byte >> 2); + dst[dst_byte_offset] = (src_byte << 6); + current_bit_offset = 2; + break; + case 2: + dst[dst_byte_offset++] |= (src_byte); + dst[dst_byte_offset] = 0; + current_bit_offset = 0; + break; + } + src_byte_offset++; + } + (*dst_size) = dst_byte_offset; +out: + return rc; +} + +/** + * ecryptfs_encrypt_and_encode_filename - converts a plaintext file name to cipher text + * @crypt_stat: The crypt_stat struct associated with the file anem to encode + * @name: The plaintext name + * @length: The length of the plaintext + * @encoded_name: The encypted name + * + * Encrypts and encodes a filename into something that constitutes a + * valid filename for a filesystem, with printable characters. + * + * We assume that we have a properly initialized crypto context, + * pointed to by crypt_stat->tfm. + * + * Returns zero on success; non-zero on otherwise + */ +int ecryptfs_encrypt_and_encode_filename( + char **encoded_name, + size_t *encoded_name_size, + struct ecryptfs_crypt_stat *crypt_stat, + struct ecryptfs_mount_crypt_stat *mount_crypt_stat, + const char *name, size_t name_size) +{ + size_t encoded_name_no_prefix_size; + int rc = 0; + + (*encoded_name) = NULL; + (*encoded_name_size) = 0; + if ((crypt_stat && (crypt_stat->flags & ECRYPTFS_ENCRYPT_FILENAMES)) + || (mount_crypt_stat && (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES))) { + struct ecryptfs_filename *filename; + + filename = kzalloc(sizeof(*filename), GFP_KERNEL); + if (!filename) { + printk(KERN_ERR "%s: Out of memory whilst attempting " + "to kzalloc [%d] bytes\n", __func__, + sizeof(*filename)); + rc = -ENOMEM; + goto out; + } + filename->filename = (char *)name; + filename->filename_size = name_size; + rc = ecryptfs_encrypt_filename(filename, crypt_stat, + mount_crypt_stat); + if (rc) { + printk(KERN_ERR "%s: Error attempting to encrypt " + "filename; rc = [%d]\n", __func__, rc); + kfree(filename); + goto out; + } + ecryptfs_encode_for_filename( + NULL, &encoded_name_no_prefix_size, + filename->encrypted_filename, + filename->encrypted_filename_size); + if ((crypt_stat && (crypt_stat->flags + & ECRYPTFS_ENCFN_USE_MOUNT_FNEK)) + || (mount_crypt_stat + && (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCFN_USE_MOUNT_FNEK))) + (*encoded_name_size) = + (ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE + + encoded_name_no_prefix_size); + else + (*encoded_name_size) = + (ECRYPTFS_FEK_ENCRYPTED_FILENAME_PREFIX_SIZE + + encoded_name_no_prefix_size); + (*encoded_name) = kmalloc((*encoded_name_size) + 1, GFP_KERNEL); + if (!(*encoded_name)) { + printk(KERN_ERR "%s: Out of memory whilst attempting " + "to kzalloc [%d] bytes\n", __func__, + (*encoded_name_size)); + rc = -ENOMEM; + kfree(filename->encrypted_filename); + kfree(filename); + goto out; + } + if ((crypt_stat && (crypt_stat->flags + & ECRYPTFS_ENCFN_USE_MOUNT_FNEK)) + || (mount_crypt_stat + && (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCFN_USE_MOUNT_FNEK))) { + memcpy((*encoded_name), + ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX, + ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE); + ecryptfs_encode_for_filename( + ((*encoded_name) + + ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE), + &encoded_name_no_prefix_size, + filename->encrypted_filename, + filename->encrypted_filename_size); + (*encoded_name_size) = + (ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE + + encoded_name_no_prefix_size); + (*encoded_name)[(*encoded_name_size)] = '\0'; + (*encoded_name_size)++; + } else { + rc = -ENOTSUPP; + } + if (rc) { + printk(KERN_ERR "%s: Error attempting to encode " + "encrypted filename; rc = [%d]\n", __func__, + rc); + kfree((*encoded_name)); + (*encoded_name) = NULL; + (*encoded_name_size) = 0; + } + kfree(filename->encrypted_filename); + kfree(filename); + } else { + rc = ecryptfs_copy_filename(encoded_name, + encoded_name_size, + name, name_size); + } +out: + return rc; +} + +/** + * ecryptfs_decode_and_decrypt_filename - converts the encoded cipher text name to decoded plaintext + * @plaintext_name: The plaintext name + * @plaintext_name_size: The plaintext name size + * @ecryptfs_dir_dentry: eCryptfs directory dentry + * @name: The filename in cipher text + * @name_size: The cipher text name size + * + * Decrypts and decodes the filename. + * + * Returns zero on error; non-zero otherwise + */ +int ecryptfs_decode_and_decrypt_filename(char **plaintext_name, + size_t *plaintext_name_size, + struct dentry *ecryptfs_dir_dentry, + const char *name, size_t name_size) +{ + char *decoded_name; + size_t decoded_name_size; + size_t packet_size; + int rc = 0; + + if ((name_size > ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE) + && (strncmp(name, ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX, + ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE) == 0)) { + struct ecryptfs_mount_crypt_stat *mount_crypt_stat = + &ecryptfs_superblock_to_private( + ecryptfs_dir_dentry->d_sb)->mount_crypt_stat; + const char *orig_name = name; + size_t orig_name_size = name_size; + + name += ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE; + name_size -= ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE; + rc = ecryptfs_decode_from_filename(NULL, &decoded_name_size, + name, name_size); + if (rc) { + printk(KERN_ERR "%s: Error attempting to decode " + "filename; rc = [%d]\n", __func__, rc); + rc = ecryptfs_copy_filename(plaintext_name, + plaintext_name_size, + orig_name, orig_name_size); + goto out; + } + decoded_name = kmalloc(decoded_name_size, GFP_KERNEL); + if (!decoded_name) { + printk(KERN_ERR "%s: Out of memory whilst attempting " + "to kmalloc [%Zd] bytes\n", __func__, + decoded_name_size); + rc = -ENOMEM; + goto out; + } + rc = ecryptfs_decode_from_filename(decoded_name, + &decoded_name_size, + name, name_size); + if (rc) { + printk(KERN_ERR "%s: Error attempting to decode " + "filename; rc = [%d]\n", __func__, rc); + rc = ecryptfs_copy_filename(plaintext_name, + plaintext_name_size, + orig_name, orig_name_size); + goto out_free; + } + rc = ecryptfs_parse_tag_70_packet(plaintext_name, + plaintext_name_size, + &packet_size, + mount_crypt_stat, + decoded_name, + decoded_name_size); + if (rc) { + printk(KERN_INFO "%s: Could not parse tag 70 packet " + "from filename; copying through filename " + "as-is\n", __func__); + rc = ecryptfs_copy_filename(plaintext_name, + plaintext_name_size, + orig_name, orig_name_size); + goto out_free; + } + } else { + rc = ecryptfs_copy_filename(plaintext_name, + plaintext_name_size, + name, name_size); + goto out; + } +out_free: + kfree(decoded_name); +out: + return rc; +} -- cgit v1.2.3-70-g09d2 From addd65ad8d19a7d7982130b16f957d5d01d3f8df Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:42:00 -0800 Subject: eCryptfs: Filename Encryption: filldir, lookup, and readlink Make the requisite modifications to ecryptfs_filldir(), ecryptfs_lookup(), and ecryptfs_readlink() to call out to filename encryption functions. Propagate filename encryption policy flags from mount-wide crypt_stat to inode crypt_stat. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 107 ++------------- fs/ecryptfs/ecryptfs_kernel.h | 6 - fs/ecryptfs/file.c | 30 ++--- fs/ecryptfs/inode.c | 294 ++++++++++++++++++++++++------------------ 4 files changed, 195 insertions(+), 242 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 18c78abba68..ea2afd2ce22 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -924,6 +924,15 @@ static void ecryptfs_copy_mount_wide_flags_to_inode_flags( crypt_stat->flags |= ECRYPTFS_METADATA_IN_XATTR; if (mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED) crypt_stat->flags |= ECRYPTFS_VIEW_AS_ENCRYPTED; + if (mount_crypt_stat->flags & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES) { + crypt_stat->flags |= ECRYPTFS_ENCRYPT_FILENAMES; + if (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCFN_USE_MOUNT_FNEK) + crypt_stat->flags |= ECRYPTFS_ENCFN_USE_MOUNT_FNEK; + else if (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCFN_USE_FEK) + crypt_stat->flags |= ECRYPTFS_ENCFN_USE_FEK; + } } static int ecryptfs_copy_mount_wide_sigs_to_inode_sigs( @@ -1060,7 +1069,8 @@ struct ecryptfs_flag_map_elem { static struct ecryptfs_flag_map_elem ecryptfs_flag_map[] = { {0x00000001, ECRYPTFS_ENABLE_HMAC}, {0x00000002, ECRYPTFS_ENCRYPTED}, - {0x00000004, ECRYPTFS_METADATA_IN_XATTR} + {0x00000004, ECRYPTFS_METADATA_IN_XATTR}, + {0x00000008, ECRYPTFS_ENCRYPT_FILENAMES} }; /** @@ -1213,6 +1223,8 @@ int ecryptfs_read_and_validate_header_region(char *data, &(ecryptfs_inode_to_private(ecryptfs_inode)->crypt_stat); int rc; + if (crypt_stat->extent_size == 0) + crypt_stat->extent_size = ECRYPTFS_DEFAULT_EXTENT_SIZE; rc = ecryptfs_read_lower(data, 0, crypt_stat->extent_size, ecryptfs_inode); if (rc) { @@ -1222,7 +1234,6 @@ int ecryptfs_read_and_validate_header_region(char *data, } if (!contains_ecryptfs_marker(data + ECRYPTFS_FILE_SIZE_BYTES)) { rc = -EINVAL; - ecryptfs_printk(KERN_DEBUG, "Valid marker not found\n"); } out: return rc; @@ -1628,98 +1639,6 @@ out: return rc; } -/** - * ecryptfs_encode_filename - converts a plaintext file name to cipher text - * @crypt_stat: The crypt_stat struct associated with the file anem to encode - * @name: The plaintext name - * @length: The length of the plaintext - * @encoded_name: The encypted name - * - * Encrypts and encodes a filename into something that constitutes a - * valid filename for a filesystem, with printable characters. - * - * We assume that we have a properly initialized crypto context, - * pointed to by crypt_stat->tfm. - * - * TODO: Implement filename decoding and decryption here, in place of - * memcpy. We are keeping the framework around for now to (1) - * facilitate testing of the components needed to implement filename - * encryption and (2) to provide a code base from which other - * developers in the community can easily implement this feature. - * - * Returns the length of encoded filename; negative if error - */ -int -ecryptfs_encode_filename(struct ecryptfs_crypt_stat *crypt_stat, - const char *name, int length, char **encoded_name) -{ - int error = 0; - - (*encoded_name) = kmalloc(length + 2, GFP_KERNEL); - if (!(*encoded_name)) { - error = -ENOMEM; - goto out; - } - /* TODO: Filename encryption is a scheduled feature for a - * future version of eCryptfs. This function is here only for - * the purpose of providing a framework for other developers - * to easily implement filename encryption. Hint: Replace this - * memcpy() with a call to encrypt and encode the - * filename, the set the length accordingly. */ - memcpy((void *)(*encoded_name), (void *)name, length); - (*encoded_name)[length] = '\0'; - error = length + 1; -out: - return error; -} - -/** - * ecryptfs_decode_filename - converts the cipher text name to plaintext - * @crypt_stat: The crypt_stat struct associated with the file - * @name: The filename in cipher text - * @length: The length of the cipher text name - * @decrypted_name: The plaintext name - * - * Decodes and decrypts the filename. - * - * We assume that we have a properly initialized crypto context, - * pointed to by crypt_stat->tfm. - * - * TODO: Implement filename decoding and decryption here, in place of - * memcpy. We are keeping the framework around for now to (1) - * facilitate testing of the components needed to implement filename - * encryption and (2) to provide a code base from which other - * developers in the community can easily implement this feature. - * - * Returns the length of decoded filename; negative if error - */ -int -ecryptfs_decode_filename(struct ecryptfs_crypt_stat *crypt_stat, - const char *name, int length, char **decrypted_name) -{ - int error = 0; - - (*decrypted_name) = kmalloc(length + 2, GFP_KERNEL); - if (!(*decrypted_name)) { - error = -ENOMEM; - goto out; - } - /* TODO: Filename encryption is a scheduled feature for a - * future version of eCryptfs. This function is here only for - * the purpose of providing a framework for other developers - * to easily implement filename encryption. Hint: Replace this - * memcpy() with a call to decode and decrypt the - * filename, the set the length accordingly. */ - memcpy((void *)(*decrypted_name), (void *)name, length); - (*decrypted_name)[length + 1] = '\0'; /* Only for convenience - * in printing out the - * string in debug - * messages */ - error = length; -out: - return error; -} - /** * ecryptfs_encrypt_filename - encrypt filename * diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index b648175a44c..c11fc95714a 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -627,12 +627,6 @@ int ecryptfs_decode_and_decrypt_filename(char **decrypted_name, struct dentry *ecryptfs_dentry, const char *name, size_t name_size); int ecryptfs_fill_zeros(struct file *file, loff_t new_length); -int ecryptfs_decode_filename(struct ecryptfs_crypt_stat *crypt_stat, - const char *name, int length, - char **decrypted_name); -int ecryptfs_encode_filename(struct ecryptfs_crypt_stat *crypt_stat, - const char *name, int length, - char **encoded_name); int ecryptfs_encrypt_and_encode_filename( char **encoded_name, size_t *encoded_name_size, diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index 71383437122..567eb4bee1b 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -77,27 +77,27 @@ struct ecryptfs_getdents_callback { /* Inspired by generic filldir in fs/readdir.c */ static int -ecryptfs_filldir(void *dirent, const char *name, int namelen, loff_t offset, - u64 ino, unsigned int d_type) +ecryptfs_filldir(void *dirent, const char *lower_name, int lower_namelen, + loff_t offset, u64 ino, unsigned int d_type) { - struct ecryptfs_crypt_stat *crypt_stat; struct ecryptfs_getdents_callback *buf = (struct ecryptfs_getdents_callback *)dirent; + int name_size; + char *name; int rc; - int decoded_length; - char *decoded_name; - crypt_stat = ecryptfs_dentry_to_private(buf->dentry)->crypt_stat; buf->filldir_called++; - decoded_length = ecryptfs_decode_filename(crypt_stat, name, namelen, - &decoded_name); - if (decoded_length < 0) { - rc = decoded_length; + rc = ecryptfs_decode_and_decrypt_filename(&name, &name_size, + buf->dentry, lower_name, + lower_namelen); + if (rc) { + printk(KERN_ERR "%s: Error attempting to decode and decrypt " + "filename [%s]; rc = [%d]\n", __func__, lower_name, + rc); goto out; } - rc = buf->filldir(buf->dirent, decoded_name, decoded_length, offset, - ino, d_type); - kfree(decoded_name); + rc = buf->filldir(buf->dirent, name, name_size, offset, ino, d_type); + kfree(name); if (rc >= 0) buf->entries_written++; out: @@ -106,8 +106,8 @@ out: /** * ecryptfs_readdir - * @file: The ecryptfs file struct - * @dirent: Directory entry + * @file: The eCryptfs directory file + * @dirent: Directory entry handle * @filldir: The filldir callback function */ static int ecryptfs_readdir(struct file *file, void *dirent, filldir_t filldir) diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 0111906a887..38309ce94d7 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -228,8 +228,7 @@ ecryptfs_create(struct inode *directory_inode, struct dentry *ecryptfs_dentry, { int rc; - /* ecryptfs_do_create() calls ecryptfs_interpose(), which opens - * the crypt_stat->lower_file (persistent file) */ + /* ecryptfs_do_create() calls ecryptfs_interpose() */ rc = ecryptfs_do_create(directory_inode, ecryptfs_dentry, mode, nd); if (unlikely(rc)) { ecryptfs_printk(KERN_WARNING, "Failed to create file in" @@ -244,141 +243,91 @@ out: } /** - * ecryptfs_lookup - * @dir: inode - * @dentry: The dentry - * @nd: nameidata, may be NULL - * - * Find a file on disk. If the file does not exist, then we'll add it to the - * dentry cache and continue on to read it from the disk. + * ecryptfs_lookup_and_interpose_lower - Perform a lookup */ -static struct dentry *ecryptfs_lookup(struct inode *dir, struct dentry *dentry, - struct nameidata *nd) +int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, + struct dentry *lower_dentry, + struct ecryptfs_crypt_stat *crypt_stat, + struct inode *ecryptfs_dir_inode, + struct nameidata *ecryptfs_nd) { - int rc = 0; struct dentry *lower_dir_dentry; - struct dentry *lower_dentry; struct vfsmount *lower_mnt; - char *encoded_name; - int encoded_namelen; - struct ecryptfs_crypt_stat *crypt_stat = NULL; + struct inode *lower_inode; struct ecryptfs_mount_crypt_stat *mount_crypt_stat; char *page_virt = NULL; - struct inode *lower_inode; u64 file_size; + int rc = 0; - lower_dir_dentry = ecryptfs_dentry_to_lower(dentry->d_parent); - dentry->d_op = &ecryptfs_dops; - if ((dentry->d_name.len == 1 && !strcmp(dentry->d_name.name, ".")) - || (dentry->d_name.len == 2 - && !strcmp(dentry->d_name.name, ".."))) { - d_drop(dentry); - goto out; - } - encoded_namelen = ecryptfs_encode_filename(crypt_stat, - dentry->d_name.name, - dentry->d_name.len, - &encoded_name); - if (encoded_namelen < 0) { - rc = encoded_namelen; - d_drop(dentry); - goto out; - } - ecryptfs_printk(KERN_DEBUG, "encoded_name = [%s]; encoded_namelen " - "= [%d]\n", encoded_name, encoded_namelen); - lower_dentry = lookup_one_len(encoded_name, lower_dir_dentry, - encoded_namelen - 1); - kfree(encoded_name); - if (IS_ERR(lower_dentry)) { - ecryptfs_printk(KERN_ERR, "ERR from lower_dentry\n"); - rc = PTR_ERR(lower_dentry); - d_drop(dentry); - goto out; - } - lower_mnt = mntget(ecryptfs_dentry_to_lower_mnt(dentry->d_parent)); - ecryptfs_printk(KERN_DEBUG, "lower_dentry = [%p]; lower_dentry->" - "d_name.name = [%s]\n", lower_dentry, - lower_dentry->d_name.name); + lower_dir_dentry = lower_dentry->d_parent; + lower_mnt = mntget(ecryptfs_dentry_to_lower_mnt( + ecryptfs_dentry->d_parent)); lower_inode = lower_dentry->d_inode; - fsstack_copy_attr_atime(dir, lower_dir_dentry->d_inode); + fsstack_copy_attr_atime(ecryptfs_dir_inode, lower_dir_dentry->d_inode); BUG_ON(!atomic_read(&lower_dentry->d_count)); - ecryptfs_set_dentry_private(dentry, + ecryptfs_set_dentry_private(ecryptfs_dentry, kmem_cache_alloc(ecryptfs_dentry_info_cache, GFP_KERNEL)); - if (!ecryptfs_dentry_to_private(dentry)) { + if (!ecryptfs_dentry_to_private(ecryptfs_dentry)) { rc = -ENOMEM; - ecryptfs_printk(KERN_ERR, "Out of memory whilst attempting " - "to allocate ecryptfs_dentry_info struct\n"); + printk(KERN_ERR "%s: Out of memory whilst attempting " + "to allocate ecryptfs_dentry_info struct\n", + __func__); goto out_dput; } - ecryptfs_set_dentry_lower(dentry, lower_dentry); - ecryptfs_set_dentry_lower_mnt(dentry, lower_mnt); + ecryptfs_set_dentry_lower(ecryptfs_dentry, lower_dentry); + ecryptfs_set_dentry_lower_mnt(ecryptfs_dentry, lower_mnt); if (!lower_dentry->d_inode) { /* We want to add because we couldn't find in lower */ - d_add(dentry, NULL); + d_add(ecryptfs_dentry, NULL); goto out; } - rc = ecryptfs_interpose(lower_dentry, dentry, dir->i_sb, - ECRYPTFS_INTERPOSE_FLAG_D_ADD); + rc = ecryptfs_interpose(lower_dentry, ecryptfs_dentry, + ecryptfs_dir_inode->i_sb, 1); if (rc) { - ecryptfs_printk(KERN_ERR, "Error interposing\n"); + printk(KERN_ERR "%s: Error interposing; rc = [%d]\n", + __func__, rc); goto out; } - if (S_ISDIR(lower_inode->i_mode)) { - ecryptfs_printk(KERN_DEBUG, "Is a directory; returning\n"); + if (S_ISDIR(lower_inode->i_mode)) goto out; - } - if (S_ISLNK(lower_inode->i_mode)) { - ecryptfs_printk(KERN_DEBUG, "Is a symlink; returning\n"); + if (S_ISLNK(lower_inode->i_mode)) goto out; - } - if (special_file(lower_inode->i_mode)) { - ecryptfs_printk(KERN_DEBUG, "Is a special file; returning\n"); + if (special_file(lower_inode->i_mode)) goto out; - } - if (!nd) { - ecryptfs_printk(KERN_DEBUG, "We have a NULL nd, just leave" - "as we *think* we are about to unlink\n"); + if (!ecryptfs_nd) goto out; - } /* Released in this function */ - page_virt = kmem_cache_zalloc(ecryptfs_header_cache_2, - GFP_USER); + page_virt = kmem_cache_zalloc(ecryptfs_header_cache_2, GFP_USER); if (!page_virt) { + printk(KERN_ERR "%s: Cannot kmem_cache_zalloc() a page\n", + __func__); rc = -ENOMEM; - ecryptfs_printk(KERN_ERR, - "Cannot ecryptfs_kmalloc a page\n"); goto out; } - crypt_stat = &ecryptfs_inode_to_private(dentry->d_inode)->crypt_stat; - if (!(crypt_stat->flags & ECRYPTFS_POLICY_APPLIED)) - ecryptfs_set_default_sizes(crypt_stat); - if (!ecryptfs_inode_to_private(dentry->d_inode)->lower_file) { - rc = ecryptfs_init_persistent_file(dentry); + if (!ecryptfs_inode_to_private(ecryptfs_dentry->d_inode)->lower_file) { + rc = ecryptfs_init_persistent_file(ecryptfs_dentry); if (rc) { printk(KERN_ERR "%s: Error attempting to initialize " "the persistent file for the dentry with name " "[%s]; rc = [%d]\n", __func__, - dentry->d_name.name, rc); - goto out; + ecryptfs_dentry->d_name.name, rc); + goto out_free_kmem; } } rc = ecryptfs_read_and_validate_header_region(page_virt, - dentry->d_inode); + ecryptfs_dentry->d_inode); if (rc) { - rc = ecryptfs_read_and_validate_xattr_region(page_virt, dentry); + rc = ecryptfs_read_and_validate_xattr_region(page_virt, + ecryptfs_dentry); if (rc) { - printk(KERN_DEBUG "Valid metadata not found in header " - "region or xattr region; treating file as " - "unencrypted\n"); rc = 0; - kmem_cache_free(ecryptfs_header_cache_2, page_virt); - goto out; + goto out_free_kmem; } crypt_stat->flags |= ECRYPTFS_METADATA_IN_XATTR; } mount_crypt_stat = &ecryptfs_superblock_to_private( - dentry->d_sb)->mount_crypt_stat; + ecryptfs_dentry->d_sb)->mount_crypt_stat; if (mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED) { if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) file_size = (crypt_stat->num_header_bytes_at_front @@ -388,14 +337,103 @@ static struct dentry *ecryptfs_lookup(struct inode *dir, struct dentry *dentry, } else { file_size = get_unaligned_be64(page_virt); } - i_size_write(dentry->d_inode, (loff_t)file_size); + i_size_write(ecryptfs_dentry->d_inode, (loff_t)file_size); +out_free_kmem: kmem_cache_free(ecryptfs_header_cache_2, page_virt); goto out; - out_dput: dput(lower_dentry); - d_drop(dentry); + d_drop(ecryptfs_dentry); out: + return rc; +} + +/** + * ecryptfs_lookup + * @ecryptfs_dir_inode: The eCryptfs directory inode + * @ecryptfs_dentry: The eCryptfs dentry that we are looking up + * @ecryptfs_nd: nameidata; may be NULL + * + * Find a file on disk. If the file does not exist, then we'll add it to the + * dentry cache and continue on to read it from the disk. + */ +static struct dentry *ecryptfs_lookup(struct inode *ecryptfs_dir_inode, + struct dentry *ecryptfs_dentry, + struct nameidata *ecryptfs_nd) +{ + char *encrypted_and_encoded_name = NULL; + int encrypted_and_encoded_name_size; + struct ecryptfs_crypt_stat *crypt_stat = NULL; + struct ecryptfs_mount_crypt_stat *mount_crypt_stat = NULL; + struct ecryptfs_inode_info *inode_info; + struct dentry *lower_dir_dentry, *lower_dentry; + int rc = 0; + + ecryptfs_dentry->d_op = &ecryptfs_dops; + if ((ecryptfs_dentry->d_name.len == 1 + && !strcmp(ecryptfs_dentry->d_name.name, ".")) + || (ecryptfs_dentry->d_name.len == 2 + && !strcmp(ecryptfs_dentry->d_name.name, ".."))) { + goto out_d_drop; + } + lower_dir_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry->d_parent); + lower_dentry = lookup_one_len(ecryptfs_dentry->d_name.name, + lower_dir_dentry, + ecryptfs_dentry->d_name.len); + if (IS_ERR(lower_dentry)) { + rc = PTR_ERR(lower_dentry); + printk(KERN_ERR "%s: lookup_one_len() returned [%d] on " + "lower_dentry = [%s]\n", __func__, rc, + ecryptfs_dentry->d_name.name); + goto out_d_drop; + } + if (lower_dentry->d_inode) + goto lookup_and_interpose; + inode_info = ecryptfs_inode_to_private(ecryptfs_dentry->d_inode); + if (inode_info) { + crypt_stat = &inode_info->crypt_stat; + /* TODO: lock for crypt_stat comparison */ + if (!(crypt_stat->flags & ECRYPTFS_POLICY_APPLIED)) + ecryptfs_set_default_sizes(crypt_stat); + } + if (crypt_stat) + mount_crypt_stat = crypt_stat->mount_crypt_stat; + else + mount_crypt_stat = &ecryptfs_superblock_to_private( + ecryptfs_dentry->d_sb)->mount_crypt_stat; + if (!(crypt_stat && (crypt_stat->flags & ECRYPTFS_ENCRYPT_FILENAMES)) + && !(mount_crypt_stat && (mount_crypt_stat->flags + & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES))) + goto lookup_and_interpose; + dput(lower_dentry); + rc = ecryptfs_encrypt_and_encode_filename( + &encrypted_and_encoded_name, &encrypted_and_encoded_name_size, + crypt_stat, mount_crypt_stat, ecryptfs_dentry->d_name.name, + ecryptfs_dentry->d_name.len); + if (rc) { + printk(KERN_ERR "%s: Error attempting to encrypt and encode " + "filename; rc = [%d]\n", __func__, rc); + goto out_d_drop; + } + lower_dentry = lookup_one_len(encrypted_and_encoded_name, + lower_dir_dentry, + encrypted_and_encoded_name_size - 1); + if (IS_ERR(lower_dentry)) { + rc = PTR_ERR(lower_dentry); + printk(KERN_ERR "%s: lookup_one_len() returned [%d] on " + "lower_dentry = [%s]\n", __func__, rc, + encrypted_and_encoded_name); + goto out_d_drop; + } +lookup_and_interpose: + rc = ecryptfs_lookup_and_interpose_lower(ecryptfs_dentry, lower_dentry, + crypt_stat, ecryptfs_dir_inode, + ecryptfs_nd); + goto out; +out_d_drop: + d_drop(ecryptfs_dentry); +out: + kfree(encrypted_and_encoded_name); return ERR_PTR(rc); } @@ -466,19 +504,21 @@ static int ecryptfs_symlink(struct inode *dir, struct dentry *dentry, struct dentry *lower_dentry; struct dentry *lower_dir_dentry; char *encoded_symname; - int encoded_symlen; - struct ecryptfs_crypt_stat *crypt_stat = NULL; + size_t encoded_symlen; + struct ecryptfs_mount_crypt_stat *mount_crypt_stat = NULL; lower_dentry = ecryptfs_dentry_to_lower(dentry); dget(lower_dentry); lower_dir_dentry = lock_parent(lower_dentry); - encoded_symlen = ecryptfs_encode_filename(crypt_stat, symname, - strlen(symname), - &encoded_symname); - if (encoded_symlen < 0) { - rc = encoded_symlen; + mount_crypt_stat = &ecryptfs_superblock_to_private( + dir->i_sb)->mount_crypt_stat; + rc = ecryptfs_encrypt_and_encode_filename(&encoded_symname, + &encoded_symlen, + NULL, + mount_crypt_stat, symname, + strlen(symname)); + if (rc) goto out_lock; - } rc = vfs_symlink(lower_dir_dentry->d_inode, lower_dentry, encoded_symname); kfree(encoded_symname); @@ -602,52 +642,54 @@ out_lock: } static int -ecryptfs_readlink(struct dentry *dentry, char __user * buf, int bufsiz) +ecryptfs_readlink(struct dentry *dentry, char __user *buf, int bufsiz) { - int rc; - struct dentry *lower_dentry; - char *decoded_name; char *lower_buf; - mm_segment_t old_fs; + struct dentry *lower_dentry; struct ecryptfs_crypt_stat *crypt_stat; + char *plaintext_name; + size_t plaintext_name_size; + mm_segment_t old_fs; + int rc; lower_dentry = ecryptfs_dentry_to_lower(dentry); if (!lower_dentry->d_inode->i_op->readlink) { rc = -EINVAL; goto out; } + crypt_stat = &ecryptfs_inode_to_private(dentry->d_inode)->crypt_stat; /* Released in this function */ lower_buf = kmalloc(bufsiz, GFP_KERNEL); if (lower_buf == NULL) { - ecryptfs_printk(KERN_ERR, "Out of memory\n"); + printk(KERN_ERR "%s: Out of memory whilst attempting to " + "kmalloc [%d] bytes\n", __func__, bufsiz); rc = -ENOMEM; goto out; } old_fs = get_fs(); set_fs(get_ds()); - ecryptfs_printk(KERN_DEBUG, "Calling readlink w/ " - "lower_dentry->d_name.name = [%s]\n", - lower_dentry->d_name.name); rc = lower_dentry->d_inode->i_op->readlink(lower_dentry, (char __user *)lower_buf, bufsiz); set_fs(old_fs); if (rc >= 0) { - crypt_stat = NULL; - rc = ecryptfs_decode_filename(crypt_stat, lower_buf, rc, - &decoded_name); - if (rc == -ENOMEM) + rc = ecryptfs_decode_and_decrypt_filename(&plaintext_name, + &plaintext_name_size, + dentry, lower_buf, + rc); + if (rc) { + printk(KERN_ERR "%s: Error attempting to decode and " + "decrypt filename; rc = [%d]\n", __func__, + rc); goto out_free_lower_buf; - if (rc > 0) { - ecryptfs_printk(KERN_DEBUG, "Copying [%d] bytes " - "to userspace: [%*s]\n", rc, - decoded_name); - if (copy_to_user(buf, decoded_name, rc)) - rc = -EFAULT; } - kfree(decoded_name); - fsstack_copy_attr_atime(dentry->d_inode, - lower_dentry->d_inode); + rc = copy_to_user(buf, plaintext_name, plaintext_name_size); + if (rc) + rc = -EFAULT; + else + rc = plaintext_name_size; + kfree(plaintext_name); + fsstack_copy_attr_atime(dentry->d_inode, lower_dentry->d_inode); } out_free_lower_buf: kfree(lower_buf); @@ -669,8 +711,6 @@ static void *ecryptfs_follow_link(struct dentry *dentry, struct nameidata *nd) } old_fs = get_fs(); set_fs(get_ds()); - ecryptfs_printk(KERN_DEBUG, "Calling readlink w/ " - "dentry->d_name.name = [%s]\n", dentry->d_name.name); rc = dentry->d_inode->i_op->readlink(dentry, (char __user *)buf, len); set_fs(old_fs); if (rc < 0) -- cgit v1.2.3-70-g09d2 From 87c94c4df0149786ad91d8a03c738a03369ee9c8 Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:42:01 -0800 Subject: eCryptfs: Filename Encryption: mount option Enable mount-wide filename encryption by providing the Filename Encryption Key (FNEK) signature as a mount option. Note that the ecryptfs-utils userspace package versions 61 or later support this option. When mounting with ecryptfs-utils version 61 or later, the mount helper will detect the availability of the passphrase-based filename encryption in the kernel (via the eCryptfs sysfs handle) and query the user interactively as to whether or not he wants to enable the feature for the mount. If the user enables filename encryption, the mount helper will then prompt for the FNEK signature that the user wishes to use, suggesting by default the signature for the mount passphrase that the user has already entered for encrypting the file contents. When not using the mount helper, the user can specify the signature for the passphrase key with the ecryptfs_fnek_sig= mount option. This key must be available in the user's keyring. The mount helper usually takes care of this step. If, however, the user is not mounting with the mount helper, then he will need to enter the passphrase key into his keyring with some other utility prior to mounting, such as ecryptfs-manager. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/main.c | 126 +++++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 99 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index fd630713c5c..789cf2e1be1 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -206,7 +206,9 @@ enum { ecryptfs_opt_sig, ecryptfs_opt_ecryptfs_sig, ecryptfs_opt_cipher, ecryptfs_opt_ecryptfs_cipher, ecryptfs_opt_ecryptfs_key_bytes, ecryptfs_opt_passthrough, ecryptfs_opt_xattr_metadata, - ecryptfs_opt_encrypted_view, ecryptfs_opt_err }; + ecryptfs_opt_encrypted_view, ecryptfs_opt_fnek_sig, + ecryptfs_opt_fn_cipher, ecryptfs_opt_fn_cipher_key_bytes, + ecryptfs_opt_err }; static const match_table_t tokens = { {ecryptfs_opt_sig, "sig=%s"}, @@ -217,6 +219,9 @@ static const match_table_t tokens = { {ecryptfs_opt_passthrough, "ecryptfs_passthrough"}, {ecryptfs_opt_xattr_metadata, "ecryptfs_xattr_metadata"}, {ecryptfs_opt_encrypted_view, "ecryptfs_encrypted_view"}, + {ecryptfs_opt_fnek_sig, "ecryptfs_fnek_sig=%s"}, + {ecryptfs_opt_fn_cipher, "ecryptfs_fn_cipher=%s"}, + {ecryptfs_opt_fn_cipher_key_bytes, "ecryptfs_fn_key_bytes=%u"}, {ecryptfs_opt_err, NULL} }; @@ -281,8 +286,11 @@ static int ecryptfs_parse_options(struct super_block *sb, char *options) int rc = 0; int sig_set = 0; int cipher_name_set = 0; + int fn_cipher_name_set = 0; int cipher_key_bytes; int cipher_key_bytes_set = 0; + int fn_cipher_key_bytes; + int fn_cipher_key_bytes_set = 0; struct ecryptfs_mount_crypt_stat *mount_crypt_stat = &ecryptfs_superblock_to_private(sb)->mount_crypt_stat; substring_t args[MAX_OPT_ARGS]; @@ -290,7 +298,12 @@ static int ecryptfs_parse_options(struct super_block *sb, char *options) char *sig_src; char *cipher_name_dst; char *cipher_name_src; + char *fn_cipher_name_dst; + char *fn_cipher_name_src; + char *fnek_dst; + char *fnek_src; char *cipher_key_bytes_src; + char *fn_cipher_key_bytes_src; if (!options) { rc = -EINVAL; @@ -322,10 +335,7 @@ static int ecryptfs_parse_options(struct super_block *sb, char *options) global_default_cipher_name; strncpy(cipher_name_dst, cipher_name_src, ECRYPTFS_MAX_CIPHER_NAME_SIZE); - ecryptfs_printk(KERN_DEBUG, - "The mount_crypt_stat " - "global_default_cipher_name set to: " - "[%s]\n", cipher_name_dst); + cipher_name_dst[ECRYPTFS_MAX_CIPHER_NAME_SIZE] = '\0'; cipher_name_set = 1; break; case ecryptfs_opt_ecryptfs_key_bytes: @@ -335,11 +345,6 @@ static int ecryptfs_parse_options(struct super_block *sb, char *options) &cipher_key_bytes_src, 0); mount_crypt_stat->global_default_cipher_key_size = cipher_key_bytes; - ecryptfs_printk(KERN_DEBUG, - "The mount_crypt_stat " - "global_default_cipher_key_size " - "set to: [%d]\n", mount_crypt_stat-> - global_default_cipher_key_size); cipher_key_bytes_set = 1; break; case ecryptfs_opt_passthrough: @@ -356,11 +361,51 @@ static int ecryptfs_parse_options(struct super_block *sb, char *options) mount_crypt_stat->flags |= ECRYPTFS_ENCRYPTED_VIEW_ENABLED; break; + case ecryptfs_opt_fnek_sig: + fnek_src = args[0].from; + fnek_dst = + mount_crypt_stat->global_default_fnek_sig; + strncpy(fnek_dst, fnek_src, ECRYPTFS_SIG_SIZE_HEX); + mount_crypt_stat->global_default_fnek_sig[ + ECRYPTFS_SIG_SIZE_HEX] = '\0'; + rc = ecryptfs_add_global_auth_tok( + mount_crypt_stat, + mount_crypt_stat->global_default_fnek_sig); + if (rc) { + printk(KERN_ERR "Error attempting to register " + "global fnek sig [%s]; rc = [%d]\n", + mount_crypt_stat->global_default_fnek_sig, + rc); + goto out; + } + mount_crypt_stat->flags |= + (ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES + | ECRYPTFS_GLOBAL_ENCFN_USE_MOUNT_FNEK); + break; + case ecryptfs_opt_fn_cipher: + fn_cipher_name_src = args[0].from; + fn_cipher_name_dst = + mount_crypt_stat->global_default_fn_cipher_name; + strncpy(fn_cipher_name_dst, fn_cipher_name_src, + ECRYPTFS_MAX_CIPHER_NAME_SIZE); + mount_crypt_stat->global_default_fn_cipher_name[ + ECRYPTFS_MAX_CIPHER_NAME_SIZE] = '\0'; + fn_cipher_name_set = 1; + break; + case ecryptfs_opt_fn_cipher_key_bytes: + fn_cipher_key_bytes_src = args[0].from; + fn_cipher_key_bytes = + (int)simple_strtol(fn_cipher_key_bytes_src, + &fn_cipher_key_bytes_src, 0); + mount_crypt_stat->global_default_fn_cipher_key_bytes = + fn_cipher_key_bytes; + fn_cipher_key_bytes_set = 1; + break; case ecryptfs_opt_err: default: - ecryptfs_printk(KERN_WARNING, - "eCryptfs: unrecognized option '%s'\n", - p); + printk(KERN_WARNING + "%s: eCryptfs: unrecognized option [%s]\n", + __func__, p); } } if (!sig_set) { @@ -374,33 +419,60 @@ static int ecryptfs_parse_options(struct super_block *sb, char *options) int cipher_name_len = strlen(ECRYPTFS_DEFAULT_CIPHER); BUG_ON(cipher_name_len >= ECRYPTFS_MAX_CIPHER_NAME_SIZE); - strcpy(mount_crypt_stat->global_default_cipher_name, ECRYPTFS_DEFAULT_CIPHER); } - if (!cipher_key_bytes_set) { + if ((mount_crypt_stat->flags & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES) + && !fn_cipher_name_set) + strcpy(mount_crypt_stat->global_default_fn_cipher_name, + mount_crypt_stat->global_default_cipher_name); + if (!cipher_key_bytes_set) mount_crypt_stat->global_default_cipher_key_size = 0; - } + if ((mount_crypt_stat->flags & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES) + && !fn_cipher_key_bytes_set) + mount_crypt_stat->global_default_fn_cipher_key_bytes = + mount_crypt_stat->global_default_cipher_key_size; mutex_lock(&key_tfm_list_mutex); if (!ecryptfs_tfm_exists(mount_crypt_stat->global_default_cipher_name, - NULL)) + NULL)) { rc = ecryptfs_add_new_key_tfm( NULL, mount_crypt_stat->global_default_cipher_name, mount_crypt_stat->global_default_cipher_key_size); - mutex_unlock(&key_tfm_list_mutex); - if (rc) { - printk(KERN_ERR "Error attempting to initialize cipher with " - "name = [%s] and key size = [%td]; rc = [%d]\n", - mount_crypt_stat->global_default_cipher_name, - mount_crypt_stat->global_default_cipher_key_size, rc); - rc = -EINVAL; - goto out; + if (rc) { + printk(KERN_ERR "Error attempting to initialize " + "cipher with name = [%s] and key size = [%td]; " + "rc = [%d]\n", + mount_crypt_stat->global_default_cipher_name, + mount_crypt_stat->global_default_cipher_key_size, + rc); + rc = -EINVAL; + mutex_unlock(&key_tfm_list_mutex); + goto out; + } } + if ((mount_crypt_stat->flags & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES) + && !ecryptfs_tfm_exists( + mount_crypt_stat->global_default_fn_cipher_name, NULL)) { + rc = ecryptfs_add_new_key_tfm( + NULL, mount_crypt_stat->global_default_fn_cipher_name, + mount_crypt_stat->global_default_fn_cipher_key_bytes); + if (rc) { + printk(KERN_ERR "Error attempting to initialize " + "cipher with name = [%s] and key size = [%td]; " + "rc = [%d]\n", + mount_crypt_stat->global_default_fn_cipher_name, + mount_crypt_stat->global_default_fn_cipher_key_bytes, + rc); + rc = -EINVAL; + mutex_unlock(&key_tfm_list_mutex); + goto out; + } + } + mutex_unlock(&key_tfm_list_mutex); rc = ecryptfs_init_global_auth_toks(mount_crypt_stat); - if (rc) { + if (rc) printk(KERN_WARNING "One or more global auth toks could not " "properly register; rc = [%d]\n", rc); - } out: return rc; } -- cgit v1.2.3-70-g09d2 From df261c52abdef147084c76ecf14473184e907547 Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:42:02 -0800 Subject: eCryptfs: Replace %Z with %z %Z is a gcc-ism. Using %z instead. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 8 ++++---- fs/ecryptfs/keystore.c | 18 +++++++++--------- fs/ecryptfs/messaging.c | 4 ++-- fs/ecryptfs/miscdev.c | 18 +++++++++--------- 4 files changed, 24 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index ea2afd2ce22..490b129311e 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1679,7 +1679,7 @@ ecryptfs_encrypt_filename(struct ecryptfs_filename *filename, kmalloc(filename->encrypted_filename_size, GFP_KERNEL); if (!filename->encrypted_filename) { printk(KERN_ERR "%s: Out of memory whilst attempting " - "to kmalloc [%Zd] bytes\n", __func__, + "to kmalloc [%zd] bytes\n", __func__, filename->encrypted_filename_size); rc = -ENOMEM; goto out; @@ -1752,7 +1752,7 @@ ecryptfs_process_key_cipher(struct crypto_blkcipher **key_tfm, *key_tfm = NULL; if (*key_size > ECRYPTFS_MAX_KEY_BYTES) { rc = -EINVAL; - printk(KERN_ERR "Requested key size is [%Zd] bytes; maximum " + printk(KERN_ERR "Requested key size is [%zd] bytes; maximum " "allowable is [%d]\n", *key_size, ECRYPTFS_MAX_KEY_BYTES); goto out; } @@ -1777,7 +1777,7 @@ ecryptfs_process_key_cipher(struct crypto_blkcipher **key_tfm, get_random_bytes(dummy_key, *key_size); rc = crypto_blkcipher_setkey(*key_tfm, dummy_key, *key_size); if (rc) { - printk(KERN_ERR "Error attempting to set key of size [%Zd] for " + printk(KERN_ERR "Error attempting to set key of size [%zd] for " "cipher [%s]; rc = [%d]\n", *key_size, cipher_name, rc); rc = -EINVAL; goto out; @@ -2221,7 +2221,7 @@ int ecryptfs_decode_and_decrypt_filename(char **plaintext_name, decoded_name = kmalloc(decoded_name_size, GFP_KERNEL); if (!decoded_name) { printk(KERN_ERR "%s: Out of memory whilst attempting " - "to kmalloc [%Zd] bytes\n", __func__, + "to kmalloc [%zd] bytes\n", __func__, decoded_name_size); rc = -ENOMEM; goto out; diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index dafceb5560e..e6a96e8f5e6 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -358,7 +358,7 @@ parse_tag_67_packet(struct ecryptfs_key_record *key_rec, /* verify that everything through the encrypted FEK size is present */ if (message_len < 4) { rc = -EIO; - printk(KERN_ERR "%s: message_len is [%Zd]; minimum acceptable " + printk(KERN_ERR "%s: message_len is [%zd]; minimum acceptable " "message length is [%d]\n", __func__, message_len, 4); goto out; } @@ -385,13 +385,13 @@ parse_tag_67_packet(struct ecryptfs_key_record *key_rec, i += data_len; if (message_len < (i + key_rec->enc_key_size)) { rc = -EIO; - printk(KERN_ERR "%s: message_len [%Zd]; max len is [%Zd]\n", + printk(KERN_ERR "%s: message_len [%zd]; max len is [%zd]\n", __func__, message_len, (i + key_rec->enc_key_size)); goto out; } if (key_rec->enc_key_size > ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES) { rc = -EIO; - printk(KERN_ERR "%s: Encrypted key_size [%Zd] larger than " + printk(KERN_ERR "%s: Encrypted key_size [%zd] larger than " "the maximum key size [%d]\n", __func__, key_rec->enc_key_size, ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES); @@ -511,7 +511,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, s = kmalloc(sizeof(*s), GFP_KERNEL); if (!s) { printk(KERN_ERR "%s: Out of memory whilst trying to kmalloc " - "[%d] bytes of kernel memory\n", __func__, sizeof(*s)); + "[%zd] bytes of kernel memory\n", __func__, sizeof(*s)); goto out; } s->desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; @@ -566,7 +566,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, GFP_KERNEL); if (!s->block_aligned_filename) { printk(KERN_ERR "%s: Out of kernel memory whilst attempting to " - "kzalloc [%Zd] bytes\n", __func__, + "kzalloc [%zd] bytes\n", __func__, s->block_aligned_filename_size); rc = -ENOMEM; goto out_unlock; @@ -721,7 +721,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, printk(KERN_ERR "%s: Error setting key for crypto context; " "rc = [%d]. s->auth_tok->token.password.session_key_" "encryption_key = [0x%p]; mount_crypt_stat->" - "global_default_fn_cipher_key_bytes = [%Zd]\n", __func__, + "global_default_fn_cipher_key_bytes = [%zd]\n", __func__, rc, s->auth_tok->token.password.session_key_encryption_key, mount_crypt_stat->global_default_fn_cipher_key_bytes); @@ -792,7 +792,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, } s->desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; if (max_packet_size < (1 + 1 + ECRYPTFS_SIG_SIZE + 1 + 1)) { - printk(KERN_WARNING "%s: max_packet_size is [%Zd]; it must be " + printk(KERN_WARNING "%s: max_packet_size is [%zd]; it must be " "at least [%d]\n", __func__, max_packet_size, (1 + 1 + ECRYPTFS_SIG_SIZE + 1 + 1)); rc = -EINVAL; @@ -909,7 +909,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, printk(KERN_ERR "%s: Error setting key for crypto context; " "rc = [%d]. s->auth_tok->token.password.session_key_" "encryption_key = [0x%p]; mount_crypt_stat->" - "global_default_fn_cipher_key_bytes = [%Zd]\n", __func__, + "global_default_fn_cipher_key_bytes = [%zd]\n", __func__, rc, s->auth_tok->token.password.session_key_encryption_key, mount_crypt_stat->global_default_fn_cipher_key_bytes); @@ -936,7 +936,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, s->i++; (*filename_size) = (s->block_aligned_filename_size - s->i); if (!((*filename_size) > 0 && (*filename_size < PATH_MAX))) { - printk(KERN_WARNING "%s: Filename size is [%Zd], which is " + printk(KERN_WARNING "%s: Filename size is [%zd], which is " "invalid\n", __func__, (*filename_size)); rc = -EINVAL; goto out_free_unlock; diff --git a/fs/ecryptfs/messaging.c b/fs/ecryptfs/messaging.c index 6913f727624..96ef51489e0 100644 --- a/fs/ecryptfs/messaging.c +++ b/fs/ecryptfs/messaging.c @@ -193,7 +193,7 @@ ecryptfs_spawn_daemon(struct ecryptfs_daemon **daemon, uid_t euid, (*daemon) = kzalloc(sizeof(**daemon), GFP_KERNEL); if (!(*daemon)) { rc = -ENOMEM; - printk(KERN_ERR "%s: Failed to allocate [%Zd] bytes of " + printk(KERN_ERR "%s: Failed to allocate [%zd] bytes of " "GFP_KERNEL memory\n", __func__, sizeof(**daemon)); goto out; } @@ -435,7 +435,7 @@ int ecryptfs_process_response(struct ecryptfs_message *msg, uid_t euid, msg_ctx->msg = kmalloc(msg_size, GFP_KERNEL); if (!msg_ctx->msg) { rc = -ENOMEM; - printk(KERN_ERR "%s: Failed to allocate [%Zd] bytes of " + printk(KERN_ERR "%s: Failed to allocate [%zd] bytes of " "GFP_KERNEL memory\n", __func__, msg_size); goto unlock; } diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index efd95a0ed1e..a67fea655f4 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -199,7 +199,7 @@ int ecryptfs_send_miscdev(char *data, size_t data_size, if (!msg_ctx->msg) { rc = -ENOMEM; printk(KERN_ERR "%s: Out of memory whilst attempting " - "to kmalloc(%Zd, GFP_KERNEL)\n", __func__, + "to kmalloc(%zd, GFP_KERNEL)\n", __func__, (sizeof(*msg_ctx->msg) + data_size)); goto out_unlock; } @@ -322,7 +322,7 @@ check_list: if (count < total_length) { rc = 0; printk(KERN_WARNING "%s: Only given user buffer of " - "size [%Zd], but we need [%Zd] to read the " + "size [%zd], but we need [%zd] to read the " "pending message\n", __func__, count, total_length); goto out_unlock_msg_ctx; } @@ -376,7 +376,7 @@ static int ecryptfs_miscdev_response(char *data, size_t data_size, if ((sizeof(*msg) + msg->data_len) != data_size) { printk(KERN_WARNING "%s: (sizeof(*msg) + msg->data_len) = " - "[%Zd]; data_size = [%Zd]. Invalid packet.\n", __func__, + "[%zd]; data_size = [%zd]. Invalid packet.\n", __func__, (sizeof(*msg) + msg->data_len), data_size); rc = -EINVAL; goto out; @@ -421,7 +421,7 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, data = kmalloc(count, GFP_KERNEL); if (!data) { printk(KERN_ERR "%s: Out of memory whilst attempting to " - "kmalloc([%Zd], GFP_KERNEL)\n", __func__, count); + "kmalloc([%zd], GFP_KERNEL)\n", __func__, count); goto out; } rc = copy_from_user(data, buf, count); @@ -436,8 +436,8 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, case ECRYPTFS_MSG_RESPONSE: if (count < (1 + 4 + 1 + sizeof(struct ecryptfs_message))) { printk(KERN_WARNING "%s: Minimum acceptable packet " - "size is [%Zd], but amount of data written is " - "only [%Zd]. Discarding response packet.\n", + "size is [%zd], but amount of data written is " + "only [%zd]. Discarding response packet.\n", __func__, (1 + 4 + 1 + sizeof(struct ecryptfs_message)), count); @@ -455,9 +455,9 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, } i += packet_size_length; if ((1 + 4 + packet_size_length + packet_size) != count) { - printk(KERN_WARNING "%s: (1 + packet_size_length([%Zd])" - " + packet_size([%Zd]))([%Zd]) != " - "count([%Zd]). Invalid packet format.\n", + printk(KERN_WARNING "%s: (1 + packet_size_length([%zd])" + " + packet_size([%zd]))([%zd]) != " + "count([%zd]). Invalid packet format.\n", __func__, packet_size_length, packet_size, (1 + packet_size_length + packet_size), count); goto out_free; -- cgit v1.2.3-70-g09d2 From a8f12864c52f8ab8520568dc97969c1749ae60bf Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:42:03 -0800 Subject: eCryptfs: Fix data types (int/size_t) Correct several format string data type specifiers. Correct filename size data types; they should be size_t rather than int when passed as parameters to some other functions (although note that the filenames will never be larger than int). Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 4 ++-- fs/ecryptfs/file.c | 2 +- fs/ecryptfs/inode.c | 2 +- fs/ecryptfs/keystore.c | 24 ++++++++++++------------ 4 files changed, 16 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 490b129311e..e935a222498 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -2093,7 +2093,7 @@ int ecryptfs_encrypt_and_encode_filename( filename = kzalloc(sizeof(*filename), GFP_KERNEL); if (!filename) { printk(KERN_ERR "%s: Out of memory whilst attempting " - "to kzalloc [%d] bytes\n", __func__, + "to kzalloc [%zd] bytes\n", __func__, sizeof(*filename)); rc = -ENOMEM; goto out; @@ -2127,7 +2127,7 @@ int ecryptfs_encrypt_and_encode_filename( (*encoded_name) = kmalloc((*encoded_name_size) + 1, GFP_KERNEL); if (!(*encoded_name)) { printk(KERN_ERR "%s: Out of memory whilst attempting " - "to kzalloc [%d] bytes\n", __func__, + "to kzalloc [%zd] bytes\n", __func__, (*encoded_name_size)); rc = -ENOMEM; kfree(filename->encrypted_filename); diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index 567eb4bee1b..9e944057001 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -82,7 +82,7 @@ ecryptfs_filldir(void *dirent, const char *lower_name, int lower_namelen, { struct ecryptfs_getdents_callback *buf = (struct ecryptfs_getdents_callback *)dirent; - int name_size; + size_t name_size; char *name; int rc; diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 38309ce94d7..7168a88cdbc 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -362,7 +362,7 @@ static struct dentry *ecryptfs_lookup(struct inode *ecryptfs_dir_inode, struct nameidata *ecryptfs_nd) { char *encrypted_and_encoded_name = NULL; - int encrypted_and_encoded_name_size; + size_t encrypted_and_encoded_name_size; struct ecryptfs_crypt_stat *crypt_stat = NULL; struct ecryptfs_mount_crypt_stat *mount_crypt_stat = NULL; struct ecryptfs_inode_info *inode_info; diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index e6a96e8f5e6..c90ca5dfc50 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -556,8 +556,8 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, goto out_unlock; } if (s->max_packet_size > (*remaining_bytes)) { - printk(KERN_WARNING "%s: Require [%d] bytes to write; only " - "[%d] available\n", __func__, s->max_packet_size, + printk(KERN_WARNING "%s: Require [%zd] bytes to write; only " + "[%zd] available\n", __func__, s->max_packet_size, (*remaining_bytes)); rc = -EINVAL; goto out_unlock; @@ -594,7 +594,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, mount_crypt_stat->global_default_fn_cipher_key_bytes); if (s->cipher_code == 0) { printk(KERN_WARNING "%s: Unable to generate code for " - "cipher [%s] with key bytes [%d]\n", __func__, + "cipher [%s] with key bytes [%zd]\n", __func__, mount_crypt_stat->global_default_fn_cipher_name, mount_crypt_stat->global_default_fn_cipher_key_bytes); rc = -EINVAL; @@ -693,7 +693,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, printk(KERN_ERR "%s: Internal error whilst attempting to " "convert filename memory to scatterlist; " "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%d]\n", __func__, rc, + "block_aligned_filename_size = [%zd]\n", __func__, rc, s->block_aligned_filename_size); goto out_release_free_unlock; } @@ -703,7 +703,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, printk(KERN_ERR "%s: Internal error whilst attempting to " "convert encrypted filename memory to scatterlist; " "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%d]\n", __func__, rc, + "block_aligned_filename_size = [%zd]\n", __func__, rc, s->block_aligned_filename_size); goto out_release_free_unlock; } @@ -787,7 +787,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, s = kmalloc(sizeof(*s), GFP_KERNEL); if (!s) { printk(KERN_ERR "%s: Out of memory whilst trying to kmalloc " - "[%d] bytes of kernel memory\n", __func__, sizeof(*s)); + "[%zd] bytes of kernel memory\n", __func__, sizeof(*s)); goto out; } s->desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; @@ -825,8 +825,8 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, - ECRYPTFS_SIG_SIZE - 1); if ((1 + s->packet_size_len + s->parsed_tag_70_packet_size) > max_packet_size) { - printk(KERN_WARNING "%s: max_packet_size is [%d]; real packet " - "size is [%d]\n", __func__, max_packet_size, + printk(KERN_WARNING "%s: max_packet_size is [%zd]; real packet " + "size is [%zd]\n", __func__, max_packet_size, (1 + s->packet_size_len + 1 + s->block_aligned_filename_size)); rc = -EINVAL; @@ -860,7 +860,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, printk(KERN_ERR "%s: Internal error whilst attempting to " "convert encrypted filename memory to scatterlist; " "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%d]\n", __func__, rc, + "block_aligned_filename_size = [%zd]\n", __func__, rc, s->block_aligned_filename_size); goto out_unlock; } @@ -869,7 +869,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, GFP_KERNEL); if (!s->decrypted_filename) { printk(KERN_ERR "%s: Out of memory whilst attempting to " - "kmalloc [%d] bytes\n", __func__, + "kmalloc [%zd] bytes\n", __func__, s->block_aligned_filename_size); rc = -ENOMEM; goto out_unlock; @@ -880,7 +880,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, printk(KERN_ERR "%s: Internal error whilst attempting to " "convert decrypted filename memory to scatterlist; " "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%d]\n", __func__, rc, + "block_aligned_filename_size = [%zd]\n", __func__, rc, s->block_aligned_filename_size); goto out_free_unlock; } @@ -944,7 +944,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, (*filename) = kmalloc(((*filename_size) + 1), GFP_KERNEL); if (!(*filename)) { printk(KERN_ERR "%s: Out of memory whilst attempting to " - "kmalloc [%d] bytes\n", __func__, + "kmalloc [%zd] bytes\n", __func__, ((*filename_size) + 1)); rc = -ENOMEM; goto out_free_unlock; -- cgit v1.2.3-70-g09d2 From 7d8bc2be51706152828164b305e969b4a8471041 Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:42:04 -0800 Subject: eCryptfs: kerneldoc for ecryptfs_parse_tag_70_packet() Kerneldoc updates for ecryptfs_parse_tag_70_packet(). Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/keystore.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'fs') diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index c90ca5dfc50..ff539420cc6 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -771,6 +771,17 @@ struct ecryptfs_parse_tag_70_packet_silly_stack { /** * parse_tag_70_packet - Parse and process FNEK-encrypted passphrase packet * @filename: This function kmalloc's the memory for the filename + * @filename_size: This function sets this to the amount of memory + * kmalloc'd for the filename + * @packet_size: This function sets this to the the number of octets + * in the packet parsed + * @mount_crypt_stat: The mount-wide cryptographic context + * @data: The memory location containing the start of the tag 70 + * packet + * @max_packet_size: The maximum legal size of the packet to be parsed + * from @data + * + * Returns zero on success; non-zero otherwise */ int ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, -- cgit v1.2.3-70-g09d2 From 71c11c378f46e42ca67c1e227646ce23bf43a8c6 Mon Sep 17 00:00:00 2001 From: Michael Halcrow Date: Tue, 6 Jan 2009 14:42:05 -0800 Subject: eCryptfs: Clean up ecryptfs_decode_from_filename() Flesh out the comments for ecryptfs_decode_from_filename(). Remove the return condition, since it is always 0. Signed-off-by: Michael Halcrow Cc: Dustin Kirkland Cc: Eric Sandeen Cc: Tyler Hicks Cc: David Kleikamp Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/crypto.c | 51 ++++++++++++++++++++++++--------------------------- 1 file changed, 24 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index e935a222498..c01e043670e 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1931,7 +1931,7 @@ static unsigned char *portable_filename_chars = ("-.0123456789ABCD" /* We could either offset on every reverse map or just pad some 0x00's * at the front here */ -static unsigned char filename_rev_map[] = { +static const unsigned char filename_rev_map[] = { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 7 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 15 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 23 */ @@ -2012,16 +2012,30 @@ out: return; } -int ecryptfs_decode_from_filename(unsigned char *dst, size_t *dst_size, - const unsigned char *src, size_t src_size) +/** + * ecryptfs_decode_from_filename + * @dst: If NULL, this function only sets @dst_size and returns. If + * non-NULL, this function decodes the encoded octets in @src + * into the memory that @dst points to. + * @dst_size: Set to the size of the decoded string. + * @src: The encoded set of octets to decode. + * @src_size: The size of the encoded set of octets to decode. + */ +static void +ecryptfs_decode_from_filename(unsigned char *dst, size_t *dst_size, + const unsigned char *src, size_t src_size) { u8 current_bit_offset = 0; size_t src_byte_offset = 0; size_t dst_byte_offset = 0; - int rc = 0; if (dst == NULL) { - /* Not exact; conservatively long */ + /* Not exact; conservatively long. Every block of 4 + * encoded characters decodes into a block of 3 + * decoded characters. This segment of code provides + * the caller with the maximum amount of allocated + * space that @dst will need to point to in a + * subsequent call. */ (*dst_size) = (((src_size + 1) * 3) / 4); goto out; } @@ -2055,7 +2069,7 @@ int ecryptfs_decode_from_filename(unsigned char *dst, size_t *dst_size, } (*dst_size) = dst_byte_offset; out: - return rc; + return; } /** @@ -2208,16 +2222,8 @@ int ecryptfs_decode_and_decrypt_filename(char **plaintext_name, name += ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE; name_size -= ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE; - rc = ecryptfs_decode_from_filename(NULL, &decoded_name_size, - name, name_size); - if (rc) { - printk(KERN_ERR "%s: Error attempting to decode " - "filename; rc = [%d]\n", __func__, rc); - rc = ecryptfs_copy_filename(plaintext_name, - plaintext_name_size, - orig_name, orig_name_size); - goto out; - } + ecryptfs_decode_from_filename(NULL, &decoded_name_size, + name, name_size); decoded_name = kmalloc(decoded_name_size, GFP_KERNEL); if (!decoded_name) { printk(KERN_ERR "%s: Out of memory whilst attempting " @@ -2226,17 +2232,8 @@ int ecryptfs_decode_and_decrypt_filename(char **plaintext_name, rc = -ENOMEM; goto out; } - rc = ecryptfs_decode_from_filename(decoded_name, - &decoded_name_size, - name, name_size); - if (rc) { - printk(KERN_ERR "%s: Error attempting to decode " - "filename; rc = [%d]\n", __func__, rc); - rc = ecryptfs_copy_filename(plaintext_name, - plaintext_name_size, - orig_name, orig_name_size); - goto out_free; - } + ecryptfs_decode_from_filename(decoded_name, &decoded_name_size, + name, name_size); rc = ecryptfs_parse_tag_70_packet(plaintext_name, plaintext_name_size, &packet_size, -- cgit v1.2.3-70-g09d2 From f70f582f0072f37790d2984647198deb3e7782a3 Mon Sep 17 00:00:00 2001 From: Qinghuang Feng Date: Tue, 6 Jan 2009 14:42:05 -0800 Subject: fs/ecryptfs/inode.c: cleanup kerneldoc Arguments lower_dentry and ecryptfs_dentry in ecryptfs_create_underlying_file() have been merged into dentry, now fix it. Signed-off-by: Qinghuang Feng Cc: Randy Dunlap Cc: Michael Halcrow Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ecryptfs/inode.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 7168a88cdbc..5697899a168 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -52,8 +52,7 @@ static void unlock_dir(struct dentry *dir) /** * ecryptfs_create_underlying_file * @lower_dir_inode: inode of the parent in the lower fs of the new file - * @lower_dentry: New file's dentry in the lower fs - * @ecryptfs_dentry: New file's dentry in ecryptfs + * @dentry: New file's dentry * @mode: The mode of the new file * @nd: nameidata of ecryptfs' parent's dentry & vfsmount * -- cgit v1.2.3-70-g09d2 From 730c9eeca9808fc2cfb506cc68c90aa330da17b0 Mon Sep 17 00:00:00 2001 From: Ian Kent Date: Tue, 6 Jan 2009 14:42:06 -0800 Subject: autofs4: improve parameter usage The parameter usage in the device node ioctl code uses arg1 and arg2 as parameter names. This patch redefines the parameter names to reflect what they actually are in an effort to make the code more readable. Signed-off-by: Ian Kent Signed-off-by: Jeff Moyer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/autofs4/dev-ioctl.c | 54 ++++++++++++++++-------------- include/linux/auto_dev-ioctl.h | 75 +++++++++++++++++++++++++++++++++++++++--- 2 files changed, 100 insertions(+), 29 deletions(-) (limited to 'fs') diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index 63b7c7afe8d..054d6d9ad9b 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -180,7 +180,7 @@ static int autofs_dev_ioctl_protover(struct file *fp, struct autofs_sb_info *sbi, struct autofs_dev_ioctl *param) { - param->arg1 = sbi->version; + param->protover.version = sbi->version; return 0; } @@ -189,7 +189,7 @@ static int autofs_dev_ioctl_protosubver(struct file *fp, struct autofs_sb_info *sbi, struct autofs_dev_ioctl *param) { - param->arg1 = sbi->sub_version; + param->protosubver.sub_version = sbi->sub_version; return 0; } @@ -335,13 +335,13 @@ static int autofs_dev_ioctl_openmount(struct file *fp, int err, fd; /* param->path has already been checked */ - if (!param->arg1) + if (!param->openmount.devid) return -EINVAL; param->ioctlfd = -1; path = param->path; - devid = param->arg1; + devid = param->openmount.devid; err = 0; fd = autofs_dev_ioctl_open_mountpoint(path, devid); @@ -373,7 +373,7 @@ static int autofs_dev_ioctl_ready(struct file *fp, { autofs_wqt_t token; - token = (autofs_wqt_t) param->arg1; + token = (autofs_wqt_t) param->ready.token; return autofs4_wait_release(sbi, token, 0); } @@ -388,8 +388,8 @@ static int autofs_dev_ioctl_fail(struct file *fp, autofs_wqt_t token; int status; - token = (autofs_wqt_t) param->arg1; - status = param->arg2 ? param->arg2 : -ENOENT; + token = (autofs_wqt_t) param->fail.token; + status = param->fail.status ? param->fail.status : -ENOENT; return autofs4_wait_release(sbi, token, status); } @@ -412,10 +412,10 @@ static int autofs_dev_ioctl_setpipefd(struct file *fp, int pipefd; int err = 0; - if (param->arg1 == -1) + if (param->setpipefd.pipefd == -1) return -EINVAL; - pipefd = param->arg1; + pipefd = param->setpipefd.pipefd; mutex_lock(&sbi->wq_mutex); if (!sbi->catatonic) { @@ -457,8 +457,8 @@ static int autofs_dev_ioctl_timeout(struct file *fp, { unsigned long timeout; - timeout = param->arg1; - param->arg1 = sbi->exp_timeout / HZ; + timeout = param->timeout.timeout; + param->timeout.timeout = sbi->exp_timeout / HZ; sbi->exp_timeout = timeout * HZ; return 0; } @@ -489,7 +489,7 @@ static int autofs_dev_ioctl_requester(struct file *fp, path = param->path; devid = sbi->sb->s_dev; - param->arg1 = param->arg2 = -1; + param->requester.uid = param->requester.gid = -1; /* Get nameidata of the parent directory */ err = path_lookup(path, LOOKUP_PARENT, &nd); @@ -505,8 +505,8 @@ static int autofs_dev_ioctl_requester(struct file *fp, err = 0; autofs4_expire_wait(nd.path.dentry); spin_lock(&sbi->fs_lock); - param->arg1 = ino->uid; - param->arg2 = ino->gid; + param->requester.uid = ino->uid; + param->requester.gid = ino->gid; spin_unlock(&sbi->fs_lock); } @@ -529,7 +529,7 @@ static int autofs_dev_ioctl_expire(struct file *fp, int err = -EAGAIN; int how; - how = param->arg1; + how = param->expire.how; mnt = fp->f_path.mnt; if (sbi->type & AUTOFS_TYPE_TRIGGER) @@ -565,9 +565,9 @@ static int autofs_dev_ioctl_askumount(struct file *fp, struct autofs_sb_info *sbi, struct autofs_dev_ioctl *param) { - param->arg1 = 0; + param->askumount.may_umount = 0; if (may_umount(fp->f_path.mnt)) - param->arg1 = 1; + param->askumount.may_umount = 1; return 0; } @@ -600,6 +600,7 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, struct nameidata nd; const char *path; unsigned int type; + unsigned int devid, magic; int err = -ENOENT; if (param->size <= sizeof(*param)) { @@ -608,10 +609,10 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, } path = param->path; - type = param->arg1; + type = param->ismountpoint.in.type; - param->arg1 = 0; - param->arg2 = 0; + param->ismountpoint.out.devid = devid = 0; + param->ismountpoint.out.magic = magic = 0; if (!fp || param->ioctlfd == -1) { if (type == AUTOFS_TYPE_ANY) { @@ -622,7 +623,7 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, goto out; sb = nd.path.dentry->d_sb; - param->arg1 = new_encode_dev(sb->s_dev); + devid = new_encode_dev(sb->s_dev); } else { struct autofs_info *ino; @@ -635,14 +636,14 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, goto out_release; ino = autofs4_dentry_ino(nd.path.dentry); - param->arg1 = autofs4_get_dev(ino->sbi); + devid = autofs4_get_dev(ino->sbi); } err = 0; if (nd.path.dentry->d_inode && nd.path.mnt->mnt_root == nd.path.dentry) { err = 1; - param->arg2 = nd.path.dentry->d_inode->i_sb->s_magic; + magic = nd.path.dentry->d_inode->i_sb->s_magic; } } else { dev_t devid = new_encode_dev(sbi->sb->s_dev); @@ -655,18 +656,21 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, if (err) goto out_release; - param->arg1 = autofs4_get_dev(sbi); + devid = autofs4_get_dev(sbi); err = have_submounts(nd.path.dentry); if (nd.path.mnt->mnt_mountpoint != nd.path.mnt->mnt_root) { if (follow_down(&nd.path.mnt, &nd.path.dentry)) { struct inode *inode = nd.path.dentry->d_inode; - param->arg2 = inode->i_sb->s_magic; + magic = inode->i_sb->s_magic; } } } + param->ismountpoint.out.devid = devid; + param->ismountpoint.out.magic = magic; + out_release: path_put(&nd.path); out: diff --git a/include/linux/auto_dev-ioctl.h b/include/linux/auto_dev-ioctl.h index f4d05ccd731..91a773993a5 100644 --- a/include/linux/auto_dev-ioctl.h +++ b/include/linux/auto_dev-ioctl.h @@ -10,6 +10,7 @@ #ifndef _LINUX_AUTO_DEV_IOCTL_H #define _LINUX_AUTO_DEV_IOCTL_H +#include #include #define AUTOFS_DEVICE_NAME "autofs" @@ -25,6 +26,60 @@ * An ioctl interface for autofs mount point control. */ +struct args_protover { + __u32 version; +}; + +struct args_protosubver { + __u32 sub_version; +}; + +struct args_openmount { + __u32 devid; +}; + +struct args_ready { + __u32 token; +}; + +struct args_fail { + __u32 token; + __s32 status; +}; + +struct args_setpipefd { + __s32 pipefd; +}; + +struct args_timeout { + __u64 timeout; +}; + +struct args_requester { + __u32 uid; + __u32 gid; +}; + +struct args_expire { + __u32 how; +}; + +struct args_askumount { + __u32 may_umount; +}; + +struct args_ismountpoint { + union { + struct args_in { + __u32 type; + } in; + struct args_out { + __u32 devid; + __u32 magic; + } out; + }; +}; + /* * All the ioctls use this structure. * When sending a path size must account for the total length @@ -39,20 +94,32 @@ struct autofs_dev_ioctl { * including this struct */ __s32 ioctlfd; /* automount command fd */ - __u32 arg1; /* Command parameters */ - __u32 arg2; + /* Command parameters */ + + union { + struct args_protover protover; + struct args_protosubver protosubver; + struct args_openmount openmount; + struct args_ready ready; + struct args_fail fail; + struct args_setpipefd setpipefd; + struct args_timeout timeout; + struct args_requester requester; + struct args_expire expire; + struct args_askumount askumount; + struct args_ismountpoint ismountpoint; + }; char path[0]; }; static inline void init_autofs_dev_ioctl(struct autofs_dev_ioctl *in) { + memset(in, 0, sizeof(struct autofs_dev_ioctl)); in->ver_major = AUTOFS_DEV_IOCTL_VERSION_MAJOR; in->ver_minor = AUTOFS_DEV_IOCTL_VERSION_MINOR; in->size = sizeof(struct autofs_dev_ioctl); in->ioctlfd = -1; - in->arg1 = 0; - in->arg2 = 0; return; } -- cgit v1.2.3-70-g09d2 From 41cfef2eb87694a8d64105c059b39f7bd6b7d4fe Mon Sep 17 00:00:00 2001 From: Ian Kent Date: Tue, 6 Jan 2009 14:42:07 -0800 Subject: autofs4: fix var shadowed by local delaration A local definition of devid in autofs_dev_ioctl_ismountpoint() shadows the fuction wide definition. Signed-off-by: Ian Kent Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/autofs4/dev-ioctl.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index 054d6d9ad9b..0566ff8db4c 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -646,17 +646,17 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, magic = nd.path.dentry->d_inode->i_sb->s_magic; } } else { - dev_t devid = new_encode_dev(sbi->sb->s_dev); + dev_t dev = autofs4_get_dev(sbi); err = path_lookup(path, LOOKUP_PARENT, &nd); if (err) goto out; - err = autofs_dev_ioctl_find_super(&nd, devid); + err = autofs_dev_ioctl_find_super(&nd, dev); if (err) goto out_release; - devid = autofs4_get_dev(sbi); + devid = dev; err = have_submounts(nd.path.dentry); -- cgit v1.2.3-70-g09d2 From a92daf6ba1f9ace8584edc8eb557a77aa7c2c71d Mon Sep 17 00:00:00 2001 From: Ian Kent Date: Tue, 6 Jan 2009 14:42:08 -0800 Subject: autofs4: make autofs type usage explicit - the type assigned at mount when no type is given is changed from 0 to AUTOFS_TYPE_INDIRECT. This was done because 0 and AUTOFS_TYPE_INDIRECT were being treated implicitly as the same type. - previously, an offset mount had it's type set to AUTOFS_TYPE_DIRECT|AUTOFS_TYPE_OFFSET but the mount control re-implementation needs to be able distinguish all three types. So this was changed to make the type setting explicit. - a type AUTOFS_TYPE_ANY was added for use by the re-implementation when checking if a given path is a mountpoint. It's not really a type as we use this to ask if a given path is a mountpoint in the autofs_dev_ioctl_ismountpoint() function. - functions to set and test the autofs mount types have been added to improve readability and make the type usage explicit. - the mount type is used from user space for the mount control re-implementtion so, for consistency, all the definitions have been moved to the user space include file include/linux/auto_fs4.h. Signed-off-by: Ian Kent Signed-off-by: Jeff Moyer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/autofs4/autofs_i.h | 2 -- fs/autofs4/dev-ioctl.c | 4 ++-- fs/autofs4/expire.c | 4 ++-- fs/autofs4/inode.c | 14 +++++------ fs/autofs4/waitq.c | 8 +++---- include/linux/auto_fs4.h | 62 ++++++++++++++++++++++++++++++++++++++++++++---- 6 files changed, 73 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h index e0f16da00e5..a76803108d0 100644 --- a/fs/autofs4/autofs_i.h +++ b/fs/autofs4/autofs_i.h @@ -25,8 +25,6 @@ #define AUTOFS_DEV_IOCTL_IOC_FIRST (AUTOFS_DEV_IOCTL_VERSION) #define AUTOFS_DEV_IOCTL_IOC_COUNT (AUTOFS_IOC_COUNT - 11) -#define AUTOFS_TYPE_TRIGGER (AUTOFS_TYPE_DIRECT|AUTOFS_TYPE_OFFSET) - #include #include #include diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index 0566ff8db4c..aa3f47293bb 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -532,7 +532,7 @@ static int autofs_dev_ioctl_expire(struct file *fp, how = param->expire.how; mnt = fp->f_path.mnt; - if (sbi->type & AUTOFS_TYPE_TRIGGER) + if (autofs_type_trigger(sbi->type)) dentry = autofs4_expire_direct(sbi->sb, mnt, sbi, how); else dentry = autofs4_expire_indirect(sbi->sb, mnt, sbi, how); @@ -615,7 +615,7 @@ static int autofs_dev_ioctl_ismountpoint(struct file *fp, param->ismountpoint.out.magic = magic = 0; if (!fp || param->ioctlfd == -1) { - if (type == AUTOFS_TYPE_ANY) { + if (autofs_type_any(type)) { struct super_block *sb; err = path_lookup(path, LOOKUP_FOLLOW, &nd); diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index 4b6fb3f628c..e3bd50776f9 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -63,7 +63,7 @@ static int autofs4_mount_busy(struct vfsmount *mnt, struct dentry *dentry) struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); /* This is an autofs submount, we can't expire it */ - if (sbi->type == AUTOFS_TYPE_INDIRECT) + if (autofs_type_indirect(sbi->type)) goto done; /* @@ -490,7 +490,7 @@ int autofs4_expire_multi(struct super_block *sb, struct vfsmount *mnt, if (arg && get_user(do_now, arg)) return -EFAULT; - if (sbi->type & AUTOFS_TYPE_TRIGGER) + if (autofs_type_trigger(sbi->type)) dentry = autofs4_expire_direct(sb, mnt, sbi, do_now); else dentry = autofs4_expire_indirect(sb, mnt, sbi, do_now); diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c index cfc23e53b6f..716e12b627b 100644 --- a/fs/autofs4/inode.c +++ b/fs/autofs4/inode.c @@ -197,9 +197,9 @@ static int autofs4_show_options(struct seq_file *m, struct vfsmount *mnt) seq_printf(m, ",minproto=%d", sbi->min_proto); seq_printf(m, ",maxproto=%d", sbi->max_proto); - if (sbi->type & AUTOFS_TYPE_OFFSET) + if (autofs_type_offset(sbi->type)) seq_printf(m, ",offset"); - else if (sbi->type & AUTOFS_TYPE_DIRECT) + else if (autofs_type_direct(sbi->type)) seq_printf(m, ",direct"); else seq_printf(m, ",indirect"); @@ -284,13 +284,13 @@ static int parse_options(char *options, int *pipefd, uid_t *uid, gid_t *gid, *maxproto = option; break; case Opt_indirect: - *type = AUTOFS_TYPE_INDIRECT; + set_autofs_type_indirect(type); break; case Opt_direct: - *type = AUTOFS_TYPE_DIRECT; + set_autofs_type_direct(type); break; case Opt_offset: - *type = AUTOFS_TYPE_OFFSET; + set_autofs_type_offset(type); break; default: return 1; @@ -338,7 +338,7 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) sbi->sb = s; sbi->version = 0; sbi->sub_version = 0; - sbi->type = AUTOFS_TYPE_INDIRECT; + set_autofs_type_indirect(&sbi->type); sbi->min_proto = 0; sbi->max_proto = 0; mutex_init(&sbi->wq_mutex); @@ -380,7 +380,7 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) } root_inode->i_fop = &autofs4_root_operations; - root_inode->i_op = sbi->type & AUTOFS_TYPE_TRIGGER ? + root_inode->i_op = autofs_type_trigger(sbi->type) ? &autofs4_direct_root_inode_operations : &autofs4_indirect_root_inode_operations; diff --git a/fs/autofs4/waitq.c b/fs/autofs4/waitq.c index e02cc8ae5eb..eeb24684590 100644 --- a/fs/autofs4/waitq.c +++ b/fs/autofs4/waitq.c @@ -337,7 +337,7 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, * is very similar for indirect mounts except only dentrys * in the root of the autofs file system may be negative. */ - if (sbi->type & AUTOFS_TYPE_TRIGGER) + if (autofs_type_trigger(sbi->type)) return -ENOENT; else if (!IS_ROOT(dentry->d_parent)) return -ENOENT; @@ -348,7 +348,7 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, return -ENOMEM; /* If this is a direct mount request create a dummy name */ - if (IS_ROOT(dentry) && sbi->type & AUTOFS_TYPE_TRIGGER) + if (IS_ROOT(dentry) && autofs_type_trigger(sbi->type)) qstr.len = sprintf(name, "%p", dentry); else { qstr.len = autofs4_getpath(sbi, dentry, &name); @@ -406,11 +406,11 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, type = autofs_ptype_expire_multi; } else { if (notify == NFY_MOUNT) - type = (sbi->type & AUTOFS_TYPE_TRIGGER) ? + type = autofs_type_trigger(sbi->type) ? autofs_ptype_missing_direct : autofs_ptype_missing_indirect; else - type = (sbi->type & AUTOFS_TYPE_TRIGGER) ? + type = autofs_type_trigger(sbi->type) ? autofs_ptype_expire_direct : autofs_ptype_expire_indirect; } diff --git a/include/linux/auto_fs4.h b/include/linux/auto_fs4.h index 2253716d4b9..55fa478bd63 100644 --- a/include/linux/auto_fs4.h +++ b/include/linux/auto_fs4.h @@ -29,10 +29,64 @@ #define AUTOFS_EXP_IMMEDIATE 1 #define AUTOFS_EXP_LEAVES 2 -#define AUTOFS_TYPE_ANY 0x0000 -#define AUTOFS_TYPE_INDIRECT 0x0001 -#define AUTOFS_TYPE_DIRECT 0x0002 -#define AUTOFS_TYPE_OFFSET 0x0004 +#define AUTOFS_TYPE_ANY 0U +#define AUTOFS_TYPE_INDIRECT 1U +#define AUTOFS_TYPE_DIRECT 2U +#define AUTOFS_TYPE_OFFSET 4U + +static inline void set_autofs_type_indirect(unsigned int *type) +{ + *type = AUTOFS_TYPE_INDIRECT; + return; +} + +static inline unsigned int autofs_type_indirect(unsigned int type) +{ + return (type == AUTOFS_TYPE_INDIRECT); +} + +static inline void set_autofs_type_direct(unsigned int *type) +{ + *type = AUTOFS_TYPE_DIRECT; + return; +} + +static inline unsigned int autofs_type_direct(unsigned int type) +{ + return (type == AUTOFS_TYPE_DIRECT); +} + +static inline void set_autofs_type_offset(unsigned int *type) +{ + *type = AUTOFS_TYPE_OFFSET; + return; +} + +static inline unsigned int autofs_type_offset(unsigned int type) +{ + return (type == AUTOFS_TYPE_OFFSET); +} + +static inline unsigned int autofs_type_trigger(unsigned int type) +{ + return (type == AUTOFS_TYPE_DIRECT || type == AUTOFS_TYPE_OFFSET); +} + +/* + * This isn't really a type as we use it to say "no type set" to + * indicate we want to search for "any" mount in the + * autofs_dev_ioctl_ismountpoint() device ioctl function. + */ +static inline void set_autofs_type_any(unsigned int *type) +{ + *type = AUTOFS_TYPE_ANY; + return; +} + +static inline unsigned int autofs_type_any(unsigned int type) +{ + return (type == AUTOFS_TYPE_ANY); +} /* Daemon notification packet types */ enum autofs_notify { -- cgit v1.2.3-70-g09d2 From bae8ec66554b27967f057a4b7888b09481ff1b8b Mon Sep 17 00:00:00 2001 From: Ian Kent Date: Tue, 6 Jan 2009 14:42:09 -0800 Subject: autofs4: fix string validation check order In function validate_dev_ioctl() we check that the string we've been sent is a valid path. The function that does this check assumes the string is NULL terminated but our NULL termination check isn't done until after this call. This patch changes the order of the check. Signed-off-by: Ian Kent Acked-by: Jeff Moyer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/autofs4/dev-ioctl.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index aa3f47293bb..025e105bffe 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -124,7 +124,7 @@ static inline void free_dev_ioctl(struct autofs_dev_ioctl *param) /* * Check sanity of parameter control fields and if a path is present - * check that it has a "/" and is terminated. + * check that it is terminated and contains at least one "/". */ static int validate_dev_ioctl(int cmd, struct autofs_dev_ioctl *param) { @@ -138,15 +138,16 @@ static int validate_dev_ioctl(int cmd, struct autofs_dev_ioctl *param) } if (param->size > sizeof(*param)) { - err = check_name(param->path); + err = invalid_str(param->path, + (void *) ((size_t) param + param->size)); if (err) { - AUTOFS_WARN("invalid path supplied for cmd(0x%08x)", - cmd); + AUTOFS_WARN( + "path string terminator missing for cmd(0x%08x)", + cmd); goto out; } - err = invalid_str(param->path, - (void *) ((size_t) param + param->size)); + err = check_name(param->path); if (err) { AUTOFS_WARN("invalid path supplied for cmd(0x%08x)", cmd); -- cgit v1.2.3-70-g09d2 From d6b54841f4ddd836c886d1e6ac381cf309ee98a3 Mon Sep 17 00:00:00 2001 From: Evgeniy Dushistov Date: Tue, 6 Jan 2009 14:42:38 -0800 Subject: minix: fix add link's wrong position calculation Fix the add link method. The oosition in the directory was calculated in wrong way - it had the incorrect shift direction. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Evgeniy Dushistov Cc: Nick Piggin Cc: [2.6.lots] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/minix/dir.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/minix/dir.c b/fs/minix/dir.c index f70433816a3..d4946c4c90e 100644 --- a/fs/minix/dir.c +++ b/fs/minix/dir.c @@ -280,7 +280,7 @@ int minix_add_link(struct dentry *dentry, struct inode *inode) return -EINVAL; got_it: - pos = (page->index >> PAGE_CACHE_SHIFT) + p - (char*)page_address(page); + pos = page_offset(page) + p - (char *)page_address(page); err = __minix_write_begin(NULL, page->mapping, pos, sbi->s_dirsize, AOP_FLAG_UNINTERRUPTIBLE, &page, NULL); if (err) -- cgit v1.2.3-70-g09d2 From 8cd3ac3aca3f2afe8570708066d64d893da468e8 Mon Sep 17 00:00:00 2001 From: WANG Cong Date: Tue, 6 Jan 2009 14:42:48 -0800 Subject: fs/exec.c: make do_coredump() void No one cares do_coredump()'s return value, and also it seems that it is also not necessary. So make it void. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: WANG Cong Cc: Alexander Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/exec.c | 4 ++-- include/linux/binfmts.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/exec.c b/fs/exec.c index 6b09d6fa4f7..71a6efe5d8b 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1686,7 +1686,7 @@ int get_dumpable(struct mm_struct *mm) return (ret >= 2) ? 2 : ret; } -int do_coredump(long signr, int exit_code, struct pt_regs * regs) +void do_coredump(long signr, int exit_code, struct pt_regs *regs) { struct core_state core_state; char corename[CORENAME_MAX_SIZE + 1]; @@ -1842,5 +1842,5 @@ fail_unlock: put_cred(cred); coredump_finish(mm); fail: - return retval; + return; } diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h index 0d0150b4901..77b4a9e4600 100644 --- a/include/linux/binfmts.h +++ b/include/linux/binfmts.h @@ -107,7 +107,7 @@ extern int setup_arg_pages(struct linux_binprm * bprm, extern int bprm_mm_init(struct linux_binprm *bprm); extern int copy_strings_kernel(int argc,char ** argv,struct linux_binprm *bprm); extern void install_exec_creds(struct linux_binprm *bprm); -extern int do_coredump(long signr, int exit_code, struct pt_regs * regs); +extern void do_coredump(long signr, int exit_code, struct pt_regs *regs); extern int set_binfmt(struct linux_binfmt *new); extern void free_bprm(struct linux_binprm *); -- cgit v1.2.3-70-g09d2 From e1f89ec95bd28b0927e76c46a7cc0927b7521c1d Mon Sep 17 00:00:00 2001 From: Eric Sesterhenn Date: Tue, 6 Jan 2009 14:43:12 -0800 Subject: bfs: add some basic sanity checks bfs_fill_super() already touches all inodes, so we can easily add some cheap sanity checks and check if the inode start and end blocks are smaller than the maximum number of blocks, the inode start block lies behind the end block or the file end offset is behind the end of the filesystem. Also check if the start of data offset in the super block fits the filesystem. The added sanity checks catch softlockup issues early when we try to sb_bread() lots of blocks in a loop in bfs_readdir() and bfs_find_entry(). In addition an oom issue in bfs_fill_super() is prevented by this when s_start is corrupted, which influences imap_len and we try to allocate a huge info->si_imap. Signed-off-by: Eric Sesterhenn Acked-by: Tigran Aivazian Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/bfs/inode.c | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) (limited to 'fs') diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c index 0ed57b5ee01..1d2bfafcad7 100644 --- a/fs/bfs/inode.c +++ b/fs/bfs/inode.c @@ -213,6 +213,9 @@ static void bfs_put_super(struct super_block *s) { struct bfs_sb_info *info = BFS_SB(s); + if (!info) + return; + brelse(info->si_sbh); mutex_destroy(&info->bfs_lock); kfree(info->si_imap); @@ -327,6 +330,7 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) unsigned i, imap_len; struct bfs_sb_info *info; long ret = -EINVAL; + unsigned long i_sblock, i_eblock, i_eoff, s_size; info = kzalloc(sizeof(*info), GFP_KERNEL); if (!info) @@ -350,6 +354,12 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) s->s_magic = BFS_MAGIC; info->si_sbh = bh; + + if (le32_to_cpu(bfs_sb->s_start) > le32_to_cpu(bfs_sb->s_end)) { + printf("Superblock is corrupted\n"); + goto out; + } + info->si_lasti = (le32_to_cpu(bfs_sb->s_start) - BFS_BSIZE) / sizeof(struct bfs_inode) + BFS_ROOT_INO - 1; @@ -397,6 +407,29 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) di = (struct bfs_inode *)bh->b_data + off; + /* test if filesystem is not corrupted */ + + i_eoff = le32_to_cpu(di->i_eoffset); + i_sblock = le32_to_cpu(di->i_sblock); + i_eblock = le32_to_cpu(di->i_eblock); + s_size = le32_to_cpu(bfs_sb->s_end); + + if (i_sblock > info->si_blocks || + i_eblock > info->si_blocks || + i_sblock > i_eblock || + i_eoff > s_size || + i_sblock * BFS_BSIZE > i_eoff) { + + printf("Inode 0x%08x corrupted\n", i); + + brelse(bh); + s->s_root = NULL; + kfree(info->si_imap); + kfree(info); + s->s_fs_info = NULL; + return -EIO; + } + if (!di->i_ino) { info->si_freei++; continue; -- cgit v1.2.3-70-g09d2 From 50682bb4de35544466c264c017030de826614367 Mon Sep 17 00:00:00 2001 From: Eric Sesterhenn Date: Tue, 6 Jan 2009 14:43:13 -0800 Subject: bfs: check that filesystem fits on the blockdevice Since all sanity checks rely on the validity of s_start which gets only checked to be smaller than s_end, we should also check if s_end is sane. Now we also try to retrieve the last block of the filesystem, which is computed by s_end. If this fails, something is bogus. Signed-off-by: Eric Sesterhenn Acked-by: Tigran Aivazian Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/bfs/inode.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'fs') diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c index 1d2bfafcad7..cc4062d12ca 100644 --- a/fs/bfs/inode.c +++ b/fs/bfs/inode.c @@ -390,6 +390,18 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) - le32_to_cpu(bfs_sb->s_start)) >> BFS_BSIZE_BITS; info->si_freei = 0; info->si_lf_eblk = 0; + + /* can we read the last block? */ + bh = sb_bread(s, info->si_blocks - 1); + if (!bh) { + printf("Last block not available: %lu\n", info->si_blocks - 1); + iput(inode); + ret = -EIO; + kfree(info->si_imap); + goto out; + } + brelse(bh); + bh = NULL; for (i = BFS_ROOT_INO; i <= info->si_lasti; i++) { struct bfs_inode *di; -- cgit v1.2.3-70-g09d2 From e4fefbac6c5bcb0388d95e83801210e7d81a071b Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Tue, 6 Jan 2009 10:08:33 +0000 Subject: GFS2: Set GFP_NOFS when allocating page on write We need to ensure that we always set GFP_NOFS in this one particular case when allocating pages for write. Reported-by: Fabio M. Di Nitto Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_address.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c index 6e4ea36c660..4ddab67867e 100644 --- a/fs/gfs2/ops_address.c +++ b/fs/gfs2/ops_address.c @@ -675,6 +675,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, goto out_trans_fail; error = -ENOMEM; + flags |= AOP_FLAG_NOFS; page = grab_cache_page_write_begin(mapping, index, flags); *pagep = page; if (unlikely(!page)) -- cgit v1.2.3-70-g09d2 From 0027ce681e3cd49fa34dab023574611f4384291d Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Tue, 6 Jan 2009 14:56:26 +0000 Subject: GFS2: LSF and LBD are now one and the same As a result of this recent patch: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=b3a6ffe16b5cc48abe7db8d04882dc45280eb693 We only need to depend on LBD. Reported-by: Fabio M. Di Nitto Signed-off-by: Steven Whitehouse --- fs/gfs2/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/gfs2/Kconfig b/fs/gfs2/Kconfig index ab2f57e3fb8..e563a644981 100644 --- a/fs/gfs2/Kconfig +++ b/fs/gfs2/Kconfig @@ -1,6 +1,6 @@ config GFS2_FS tristate "GFS2 file system support" - depends on EXPERIMENTAL && (64BIT || (LSF && LBD)) + depends on EXPERIMENTAL && (64BIT || LBD) select FS_POSIX_ACL select CRC32 help -- cgit v1.2.3-70-g09d2 From c8f554b947e80a90e1b43bbd4fd26c27765b5f96 Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Tue, 6 Jan 2009 10:47:50 -0600 Subject: GFS2: Fix typo in gfs_page_mkwrite() There is a typo in gfs2_page_mkwrite() gfs2_write_alloc_required() expects pos to be the offset in bytes. However, instead of the page index being shifted by by PAGE_CACHE_SHIFT, it was shifted by (PAGE_CACHE_SIZE - inode->i_blkbits). This patch simply shifts the page index by the proper amount. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index 289c5f54ba5..93fe41b67f9 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -342,7 +342,7 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct page *page) struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); unsigned long last_index; - u64 pos = page->index << (PAGE_CACHE_SIZE - inode->i_blkbits); + u64 pos = page->index << PAGE_CACHE_SHIFT; unsigned int data_blocks, ind_blocks, rblocks; int alloc_required = 0; struct gfs2_holder gh; -- cgit v1.2.3-70-g09d2 From efaee192063a54749c56b7383803e16fe553630e Mon Sep 17 00:00:00 2001 From: Arjan van de Ven Date: Tue, 6 Jan 2009 07:20:54 -0800 Subject: async: make the final inode deletion an asynchronous event this makes "rm -rf" on a (names cached) kernel tree go from 11.6 to 8.6 seconds on an ext3 filesystem Signed-off-by: Arjan van de Ven --- fs/inode.c | 20 +++++++++++++------- fs/super.c | 10 ++++++++++ include/linux/fs.h | 5 +++++ 3 files changed, 28 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index 7a6e8c2ff7b..0013ac1af8e 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -22,6 +22,7 @@ #include #include #include +#include /* * This is needed for the following functions: @@ -1138,16 +1139,11 @@ EXPORT_SYMBOL(remove_inode_hash); * I_FREEING is set so that no-one will take a new reference to the inode while * it is being deleted. */ -void generic_delete_inode(struct inode *inode) +static void generic_delete_inode_async(void *data, async_cookie_t cookie) { + struct inode *inode = data; const struct super_operations *op = inode->i_sb->s_op; - list_del_init(&inode->i_list); - list_del_init(&inode->i_sb_list); - inode->i_state |= I_FREEING; - inodes_stat.nr_inodes--; - spin_unlock(&inode_lock); - security_inode_delete(inode); if (op->delete_inode) { @@ -1171,6 +1167,16 @@ void generic_delete_inode(struct inode *inode) destroy_inode(inode); } +void generic_delete_inode(struct inode *inode) +{ + list_del_init(&inode->i_list); + list_del_init(&inode->i_sb_list); + inode->i_state |= I_FREEING; + inodes_stat.nr_inodes--; + spin_unlock(&inode_lock); + async_schedule_special(generic_delete_inode_async, inode, &inode->i_sb->s_async_list); +} + EXPORT_SYMBOL(generic_delete_inode); static void generic_forget_inode(struct inode *inode) diff --git a/fs/super.c b/fs/super.c index ddba069d7a9..cb20744ec78 100644 --- a/fs/super.c +++ b/fs/super.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include "internal.h" @@ -71,6 +72,7 @@ static struct super_block *alloc_super(struct file_system_type *type) INIT_HLIST_HEAD(&s->s_anon); INIT_LIST_HEAD(&s->s_inodes); INIT_LIST_HEAD(&s->s_dentry_lru); + INIT_LIST_HEAD(&s->s_async_list); init_rwsem(&s->s_umount); mutex_init(&s->s_lock); lockdep_set_class(&s->s_umount, &type->s_umount_key); @@ -289,11 +291,18 @@ void generic_shutdown_super(struct super_block *sb) { const struct super_operations *sop = sb->s_op; + if (sb->s_root) { shrink_dcache_for_umount(sb); fsync_super(sb); lock_super(sb); sb->s_flags &= ~MS_ACTIVE; + + /* + * wait for asynchronous fs operations to finish before going further + */ + async_synchronize_full_special(&sb->s_async_list); + /* bad name - it should be evict_inodes() */ invalidate_inodes(sb); lock_kernel(); @@ -449,6 +458,7 @@ void sync_filesystems(int wait) if (sb->s_flags & MS_RDONLY) continue; sb->s_need_sync_fs = 1; + async_synchronize_full_special(&sb->s_async_list); } restart: diff --git a/include/linux/fs.h b/include/linux/fs.h index d7eba77f666..e38a64d71ef 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1184,6 +1184,11 @@ struct super_block { * generic_show_options() */ char *s_options; + + /* + * storage for asynchronous operations + */ + struct list_head s_async_list; }; extern struct timespec current_fs_time(struct super_block *sb); -- cgit v1.2.3-70-g09d2 From 5e07878787ad07361571150230cc3a8d522ae046 Mon Sep 17 00:00:00 2001 From: Inaky Perez-Gonzalez Date: Sat, 20 Dec 2008 16:57:39 -0800 Subject: debugfs: add helpers for exporting a size_t simple value MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the same spirit as debugfs_create_*(), introduce helpers for exporting size_t values over debugfs. The only trick done is that the format verifier is kept at %llu instead of %zu; otherwise type warnings would pop up: format ‘%zu’ expects type ‘size_t’, but argument 2 has type ‘long long unsigned int’ There is no real way to fix this one--however, we can consider %llu and %zu to be compatible if we consider that we are using the same for validating in debugfs_create_{x,u}{8,16,32}(). Signed-off-by: Inaky Perez-Gonzalez Signed-off-by: Greg Kroah-Hartman --- fs/debugfs/file.c | 32 ++++++++++++++++++++++++++++++++ include/linux/debugfs.h | 2 ++ 2 files changed, 34 insertions(+) (limited to 'fs') diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 159a5efd6a8..33a90120f6a 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -294,6 +294,38 @@ struct dentry *debugfs_create_x32(const char *name, mode_t mode, } EXPORT_SYMBOL_GPL(debugfs_create_x32); + +static int debugfs_size_t_set(void *data, u64 val) +{ + *(size_t *)data = val; + return 0; +} +static int debugfs_size_t_get(void *data, u64 *val) +{ + *val = *(size_t *)data; + return 0; +} +DEFINE_SIMPLE_ATTRIBUTE(fops_size_t, debugfs_size_t_get, debugfs_size_t_set, + "%llu\n"); /* %llu and %zu are more or less the same */ + +/** + * debugfs_create_size_t - create a debugfs file that is used to read and write an size_t value + * @name: a pointer to a string containing the name of the file to create. + * @mode: the permission that the file should have + * @parent: a pointer to the parent dentry for this file. This should be a + * directory dentry if set. If this parameter is %NULL, then the + * file will be created in the root of the debugfs filesystem. + * @value: a pointer to the variable that the file should read to and write + * from. + */ +struct dentry *debugfs_create_size_t(const char *name, mode_t mode, + struct dentry *parent, size_t *value) +{ + return debugfs_create_file(name, mode, parent, value, &fops_size_t); +} +EXPORT_SYMBOL_GPL(debugfs_create_size_t); + + static ssize_t read_file_bool(struct file *file, char __user *user_buf, size_t count, loff_t *ppos) { diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h index e1a6c046cea..23936b16426 100644 --- a/include/linux/debugfs.h +++ b/include/linux/debugfs.h @@ -63,6 +63,8 @@ struct dentry *debugfs_create_x16(const char *name, mode_t mode, struct dentry *parent, u16 *value); struct dentry *debugfs_create_x32(const char *name, mode_t mode, struct dentry *parent, u32 *value); +struct dentry *debugfs_create_size_t(const char *name, mode_t mode, + struct dentry *parent, size_t *value); struct dentry *debugfs_create_bool(const char *name, mode_t mode, struct dentry *parent, u32 *value); -- cgit v1.2.3-70-g09d2 From 55ef1274dddd4de387c54d110e354ffbb6cdc706 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sat, 20 Dec 2008 11:58:38 -0800 Subject: nfsd: Ensure nfsv4 calls the underlying filesystem on LOCKT Since nfsv4 allows LOCKT without an open, but the ->lock() method is a file method, we fake up a struct file in the nfsv4 code with just the fields we need initialized. But we forgot to initialize the file operations, with the result that LOCKT never results in a call to the filesystem's ->lock() method (if it exists). We could just add that one more initialization. But this hack of faking up a struct file with only some fields initialized seems the kind of thing that might cause more problems in the future. We should either do an open and get a real struct file, or make lock-testing an inode (not a file) method. This patch does the former. Reported-by: Marc Eshel Tested-by: Marc Eshel Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 30 ++++++++++++++++++++---------- 1 file changed, 20 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 06b89df9221..e62d0e3df8b 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2781,6 +2781,25 @@ out: return status; } +/* + * The NFSv4 spec allows a client to do a LOCKT without holding an OPEN, + * so we do a temporary open here just to get an open file to pass to + * vfs_test_lock. (Arguably perhaps test_lock should be done with an + * inode operation.) + */ +static int nfsd_test_lock(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file_lock *lock) +{ + struct file *file; + int err; + + err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file); + if (err) + return err; + err = vfs_test_lock(file, lock); + nfsd_close(file); + return err; +} + /* * LOCKT operation */ @@ -2789,7 +2808,6 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, struct nfsd4_lockt *lockt) { struct inode *inode; - struct file file; struct file_lock file_lock; int error; __be32 status; @@ -2847,16 +2865,8 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nfs4_transform_lock_offset(&file_lock); - /* vfs_test_lock uses the struct file _only_ to resolve the inode. - * since LOCKT doesn't require an OPEN, and therefore a struct - * file may not exist, pass vfs_test_lock a struct file with - * only the dentry:inode set. - */ - memset(&file, 0, sizeof (struct file)); - file.f_path.dentry = cstate->current_fh.fh_dentry; - status = nfs_ok; - error = vfs_test_lock(&file, &file_lock); + error = nfsd_test_lock(rqstp, &cstate->current_fh, &file_lock); if (error) { status = nfserrno(error); goto out; -- cgit v1.2.3-70-g09d2 From d3fe5ea7cf815c037c90b1f1464ffc1ab5e8601b Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 31 Dec 2008 16:06:04 -0500 Subject: NLM: Refactor make_socks() function Clean up: extract common logic in NLM's make_socks() function into a helper. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/svc.c | 31 ++++++++++++++++--------------- 1 file changed, 16 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index 3e5f9f07911..cf3899aec37 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -204,6 +204,19 @@ lockd(void *vrqstp) return 0; } +static int create_lockd_listener(struct svc_serv *serv, char *name, + unsigned short port) +{ + struct svc_xprt *xprt; + + xprt = svc_find_xprt(serv, name, 0, 0); + if (xprt == NULL) + return svc_create_xprt(serv, name, port, SVC_SOCK_DEFAULTS); + + svc_xprt_put(xprt); + return 0; +} + /* * Ensure there are active UDP and TCP listeners for lockd. * @@ -217,23 +230,11 @@ lockd(void *vrqstp) static int make_socks(struct svc_serv *serv) { static int warned; - struct svc_xprt *xprt; int err = 0; - xprt = svc_find_xprt(serv, "udp", 0, 0); - if (!xprt) - err = svc_create_xprt(serv, "udp", nlm_udpport, - SVC_SOCK_DEFAULTS); - else - svc_xprt_put(xprt); - if (err >= 0) { - xprt = svc_find_xprt(serv, "tcp", 0, 0); - if (!xprt) - err = svc_create_xprt(serv, "tcp", nlm_tcpport, - SVC_SOCK_DEFAULTS); - else - svc_xprt_put(xprt); - } + err = create_lockd_listener(serv, "udp", nlm_udpport); + if (err >= 0) + err = create_lockd_listener(serv, "tcp", nlm_tcpport); if (err >= 0) { warned = 0; err = 0; -- cgit v1.2.3-70-g09d2 From 0dba7c2a9ed3d4a1e58f5d94fffa9f44dbe012e6 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 31 Dec 2008 16:06:11 -0500 Subject: NLM: Clean up flow of control in make_socks() function Clean up: Use Bruce's preferred control flow style in make_socks(). Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/lockd/svc.c | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index cf3899aec37..64f1c31b585 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -230,17 +230,23 @@ static int create_lockd_listener(struct svc_serv *serv, char *name, static int make_socks(struct svc_serv *serv) { static int warned; - int err = 0; + int err; err = create_lockd_listener(serv, "udp", nlm_udpport); - if (err >= 0) - err = create_lockd_listener(serv, "tcp", nlm_tcpport); - if (err >= 0) { - warned = 0; - err = 0; - } else if (warned++ == 0) + if (err < 0) + goto out_err; + + err = create_lockd_listener(serv, "tcp", nlm_tcpport); + if (err < 0) + goto out_err; + + warned = 0; + return 0; + +out_err: + if (warned++ == 0) printk(KERN_WARNING - "lockd_up: makesock failed, error=%d\n", err); + "lockd_up: makesock failed, error=%d\n", err); return err; } -- cgit v1.2.3-70-g09d2 From f05ef8db1abe68e3f6fc272efee51bc54ce528c5 Mon Sep 17 00:00:00 2001 From: David Howells Date: Mon, 5 Jan 2009 17:19:37 +0000 Subject: CRED: Fix NFSD regression Fix a regression in NFSD's permission checking introduced by the credentials patches. There are two parts to the problem, both in nfsd_setuser(): (1) The return value of set_groups() is -ve if in error, not 0, and should be checked appropriately. 0 indicates success. (2) The UID to use for fs accesses is in new->fsuid, not new->uid (which is 0). This causes CAP_DAC_OVERRIDE to always be set, rather than being cleared if the UID is anything other than 0 after squashing. Reported-by: J. Bruce Fields Signed-off-by: David Howells Acked-by: James Morris Signed-off-by: J. Bruce Fields --- fs/nfsd/auth.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/auth.c b/fs/nfsd/auth.c index 0184fe9b514..c903e04aa21 100644 --- a/fs/nfsd/auth.c +++ b/fs/nfsd/auth.c @@ -76,10 +76,10 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) ret = set_groups(new, gi); put_group_info(gi); - if (!ret) + if (ret < 0) goto error; - if (new->uid) + if (new->fsuid) new->cap_effective = cap_drop_nfsd_set(new->cap_effective); else new->cap_effective = cap_raise_nfsd_set(new->cap_effective, -- cgit v1.2.3-70-g09d2 From 9a8d248e2d2e9c880ac4561f27fea5dc200655bd Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Tue, 6 Jan 2009 13:37:03 -0500 Subject: nfsd: fix double-locks of directory mutex A number of nfsd operations depend on the i_mutex to cover more code than just the fsync, so the approach of 4c728ef583b3d8 "add a vfs_fsync helper" doesn't work for nfsd. Revert the parts of those patches that touch nfsd. Note: we can't, however, remove the logic from vfs_fsync that was needed only for the special case of nfsd, because a vfs_fsync(NULL,...) call can still result indirectly from a stackable filesystem that was called by nfsd. (Thanks to Christoph Hellwig for pointing this out.) Reported-by: Eric Sesterhenn Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 34 +++++++++++++++++++++++++++++++--- 1 file changed, 31 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 44aa92aba89..6e50aaa56ca 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -744,16 +744,44 @@ nfsd_close(struct file *filp) fput(filp); } +/* + * Sync a file + * As this calls fsync (not fdatasync) there is no need for a write_inode + * after it. + */ +static inline int nfsd_dosync(struct file *filp, struct dentry *dp, + const struct file_operations *fop) +{ + struct inode *inode = dp->d_inode; + int (*fsync) (struct file *, struct dentry *, int); + int err; + + err = filemap_fdatawrite(inode->i_mapping); + if (err == 0 && fop && (fsync = fop->fsync)) + err = fsync(filp, dp, 0); + if (err == 0) + err = filemap_fdatawait(inode->i_mapping); + + return err; +} + static int nfsd_sync(struct file *filp) { - return vfs_fsync(filp, filp->f_path.dentry, 0); + int err; + struct inode *inode = filp->f_path.dentry->d_inode; + dprintk("nfsd: sync file %s\n", filp->f_path.dentry->d_name.name); + mutex_lock(&inode->i_mutex); + err=nfsd_dosync(filp, filp->f_path.dentry, filp->f_op); + mutex_unlock(&inode->i_mutex); + + return err; } int -nfsd_sync_dir(struct dentry *dentry) +nfsd_sync_dir(struct dentry *dp) { - return vfs_fsync(NULL, dentry, 0); + return nfsd_dosync(NULL, dp, dp->d_inode->i_fop); } /* -- cgit v1.2.3-70-g09d2 From 30fa8c0157e4591ee2227aaa0b17cd3b0da5e6cb Mon Sep 17 00:00:00 2001 From: Steve Dickson Date: Wed, 7 Jan 2009 16:54:30 -0500 Subject: NFSD: FIDs need to take precedence over UUIDs When determining the fsid_type in fh_compose(), the setting of the FID via fsid= export option needs to take precedence over using the UUID device id. Signed-off-by: Steve Dickson Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsfh.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index 019a8a20184..9f1ca17293d 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -484,6 +484,8 @@ fh_compose(struct svc_fh *fhp, struct svc_export *exp, struct dentry *dentry, goto retry; break; } + } else if (exp->ex_flags & NFSEXP_FSID) { + fsid_type = FSID_NUM; } else if (exp->ex_uuid) { if (fhp->fh_maxsize >= 64) { if (root_export) @@ -496,9 +498,7 @@ fh_compose(struct svc_fh *fhp, struct svc_export *exp, struct dentry *dentry, else fsid_type = FSID_UUID4_INUM; } - } else if (exp->ex_flags & NFSEXP_FSID) - fsid_type = FSID_NUM; - else if (!old_valid_dev(ex_dev)) + } else if (!old_valid_dev(ex_dev)) /* for newer device numbers, we must use a newer fsid format */ fsid_type = FSID_ENCODE_DEV; else -- cgit v1.2.3-70-g09d2 From b7aeda40d3010666d2c024c80557b6aa92a1a1ad Mon Sep 17 00:00:00 2001 From: Dean Hildebrand Date: Mon, 15 Dec 2008 19:40:15 +0200 Subject: nfsd: add etoosmall to nfserrno Signed-off-by: Dean Hildebrand Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsproc.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c index 5cffeca7ace..6f7f2635122 100644 --- a/fs/nfsd/nfsproc.c +++ b/fs/nfsd/nfsproc.c @@ -622,6 +622,7 @@ nfserrno (int errno) { nfserr_badname, -ESRCH }, { nfserr_io, -ETXTBSY }, { nfserr_notsupp, -EOPNOTSUPP }, + { nfserr_toosmall, -ETOOSMALL }, }; int i; -- cgit v1.2.3-70-g09d2 From 0407717d8587f60003f4904bff27650cd836c00c Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Mon, 15 Dec 2008 19:40:49 +0200 Subject: nfsd: dprint each op status in nfsd4_proc_compound Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 669461e291a..9fa60a3ad48 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -946,6 +946,11 @@ encode_op: nfsd4_encode_operation(resp, op); status = op->status; } + + dprintk("nfsv4 compound op %p opcnt %d #%d: %d: status %d\n", + args->ops, args->opcnt, resp->opcnt, op->opnum, + be32_to_cpu(status)); + if (cstate->replay_owner) { nfs4_put_stateowner(cstate->replay_owner); cstate->replay_owner = NULL; -- cgit v1.2.3-70-g09d2 From df96fcf02a5fd2ae4e9b09e079dd6ef12d10ecd7 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Mon, 15 Dec 2008 19:41:10 +0200 Subject: nfsd: git rid of nfs4_cb_null_ops declaration There's no use for nfs4_cb_null_ops's declaration in fs/nfsd/nfs4callback.c Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 6d7d8c02c19..c464181b599 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -53,9 +53,6 @@ #define NFSPROC4_CB_NULL 0 #define NFSPROC4_CB_COMPOUND 1 -/* declarations */ -static const struct rpc_call_ops nfs4_cb_null_ops; - /* Index of predefined Linux callback client operations */ enum { -- cgit v1.2.3-70-g09d2 From 4e65ebf08951326709817e654c149d0a94982e01 Mon Sep 17 00:00:00 2001 From: Marc Eshel Date: Mon, 15 Dec 2008 19:41:31 +0200 Subject: nfsd: delete wrong file comment from nfsd/nfs4xdr.c Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index afcdf4b7684..f65953be39c 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1,6 +1,4 @@ /* - * fs/nfs/nfs4xdr.c - * * Server-side XDR for NFSv4 * * Copyright (c) 2002 The Regents of the University of Michigan. -- cgit v1.2.3-70-g09d2 From 87df4de8073f922a1f643b9fa6ba0412d5529ecf Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Mon, 15 Dec 2008 19:42:03 +0200 Subject: nfsd: last_byte_offset refactor the nfs4 server lock code to use last_byte_offset to compute the last byte covered by the lock. Check for overflow so that the last byte is set to NFS4_MAX_UINT64 if offset + len wraps around. Also, use NFS4_MAX_UINT64 for ~(u64)0 where appropriate. Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 42 ++++++++++++++++++++++++++---------------- include/linux/nfs4.h | 2 ++ 2 files changed, 28 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index e62d0e3df8b..88db7d3ec12 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2416,6 +2416,26 @@ out: #define LOCK_HASH_SIZE (1 << LOCK_HASH_BITS) #define LOCK_HASH_MASK (LOCK_HASH_SIZE - 1) +static inline u64 +end_offset(u64 start, u64 len) +{ + u64 end; + + end = start + len; + return end >= start ? end: NFS4_MAX_UINT64; +} + +/* last octet in a range */ +static inline u64 +last_byte_offset(u64 start, u64 len) +{ + u64 end; + + BUG_ON(!len); + end = start + len; + return end > start ? end - 1: NFS4_MAX_UINT64; +} + #define lockownerid_hashval(id) \ ((id) & LOCK_HASH_MASK) @@ -2519,8 +2539,8 @@ nfs4_set_lock_denied(struct file_lock *fl, struct nfsd4_lock_denied *deny) deny->ld_clientid.cl_id = 0; } deny->ld_start = fl->fl_start; - deny->ld_length = ~(u64)0; - if (fl->fl_end != ~(u64)0) + deny->ld_length = NFS4_MAX_UINT64; + if (fl->fl_end != NFS4_MAX_UINT64) deny->ld_length = fl->fl_end - fl->fl_start + 1; deny->ld_type = NFS4_READ_LT; if (fl->fl_type != F_RDLCK) @@ -2617,7 +2637,7 @@ out: static int check_lock_length(u64 offset, u64 length) { - return ((length == 0) || ((length != ~(u64)0) && + return ((length == 0) || ((length != NFS4_MAX_UINT64) && LOFF_OVERFLOW(offset, length))); } @@ -2737,11 +2757,7 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, file_lock.fl_lmops = &nfsd_posix_mng_ops; file_lock.fl_start = lock->lk_offset; - if ((lock->lk_length == ~(u64)0) || - LOFF_OVERFLOW(lock->lk_offset, lock->lk_length)) - file_lock.fl_end = ~(u64)0; - else - file_lock.fl_end = lock->lk_offset + lock->lk_length - 1; + file_lock.fl_end = last_byte_offset(lock->lk_offset, lock->lk_length); nfs4_transform_lock_offset(&file_lock); /* @@ -2858,10 +2874,7 @@ nfsd4_lockt(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, file_lock.fl_lmops = &nfsd_posix_mng_ops; file_lock.fl_start = lockt->lt_offset; - if ((lockt->lt_length == ~(u64)0) || LOFF_OVERFLOW(lockt->lt_offset, lockt->lt_length)) - file_lock.fl_end = ~(u64)0; - else - file_lock.fl_end = lockt->lt_offset + lockt->lt_length - 1; + file_lock.fl_end = last_byte_offset(lockt->lt_offset, lockt->lt_length); nfs4_transform_lock_offset(&file_lock); @@ -2917,10 +2930,7 @@ nfsd4_locku(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, file_lock.fl_lmops = &nfsd_posix_mng_ops; file_lock.fl_start = locku->lu_offset; - if ((locku->lu_length == ~(u64)0) || LOFF_OVERFLOW(locku->lu_offset, locku->lu_length)) - file_lock.fl_end = ~(u64)0; - else - file_lock.fl_end = locku->lu_offset + locku->lu_length - 1; + file_lock.fl_end = last_byte_offset(locku->lu_offset, locku->lu_length); nfs4_transform_lock_offset(&file_lock); /* diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index ea036676948..b912311a56b 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -88,6 +88,8 @@ #define NFS4_ACE_GENERIC_EXECUTE 0x001200A0 #define NFS4_ACE_MASK_ALL 0x001F01FF +#define NFS4_MAX_UINT64 (~(u64)0) + enum nfs4_acl_whotype { NFS4_ACL_WHO_NAMED = 0, NFS4_ACL_WHO_OWNER, -- cgit v1.2.3-70-g09d2