From patchwork Thu Aug 20 22:48:18 2009 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andrew Morton X-Patchwork-Id: 53734 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from lists.samba.org (fn.samba.org [216.83.154.106]) by ozlabs.org (Postfix) with ESMTP id 4B63BB7D1A for ; Thu, 27 May 2010 22:34:38 +1000 (EST) Received: from fn.samba.org (localhost [127.0.0.1]) by lists.samba.org (Postfix) with ESMTP id 341DEACF5E; Thu, 27 May 2010 06:34:35 -0600 (MDT) X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on fn.samba.org X-Spam-Level: X-Spam-Status: No, score=-4.1 required=3.8 tests=BAYES_00,SPF_PASS autolearn=ham version=3.2.5 X-Original-To: linux-cifs-client@lists.samba.org Delivered-To: linux-cifs-client@lists.samba.org Received: from smtp1.linux-foundation.org (smtp1.linux-foundation.org [140.211.169.13]) by lists.samba.org (Postfix) with ESMTP id 9207DAD021 for ; Thu, 20 Aug 2009 16:41:38 -0600 (MDT) Received: from imap1.linux-foundation.org (imap1.linux-foundation.org [140.211.169.55]) by smtp1.linux-foundation.org (8.14.2/8.13.5/Debian-3ubuntu1.1) with ESMTP id n7KMmJGh029928 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 20 Aug 2009 15:48:20 -0700 Received: from localhost.localdomain (localhost [127.0.0.1]) by imap1.linux-foundation.org (8.13.5.20060308/8.13.5/Debian-3ubuntu1.1) with ESMTP id n7KMmJlk029335; Thu, 20 Aug 2009 15:48:19 -0700 Message-Id: <200908202248.n7KMmJlk029335@imap1.linux-foundation.org> To: mm-commits@vger.kernel.org From: akpm@linux-foundation.org Date: Thu, 20 Aug 2009 15:48:18 -0700 X-MIMEDefang-Filter: lf$Revision: 1.188 $ X-Scanned-By: MIMEDefang 2.63 on 140.211.169.13 X-Mailman-Approved-At: Thu, 27 May 2010 06:34:18 -0600 Cc: npiggin@suse.de, linux-nfs@vger.kernel.org, jack@suse.cz, miklos@szeredi.hu, trond.myklebust@fys.uio.no, hugh.dickins@tiscali.co.uk, chris.mason@oracle.com, linux-cifs-client@lists.samba.org, hch@lst.de, hirofumi@mail.parknet.co.jp, viro@zeniv.linux.org.uk Subject: [linux-cifs-client] + fs-use-new-truncate-helpers.patch added to -mm tree X-BeenThere: linux-cifs-client@lists.samba.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: The Linux CIFS VFS client List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: linux-cifs-client-bounces@lists.samba.org Errors-To: linux-cifs-client-bounces@lists.samba.org The patch titled fs: use new truncate helpers has been added to the -mm tree. Its filename is fs-use-new-truncate-helpers.patch Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's *** Remember to use Documentation/SubmitChecklist when testing your code *** See http://userweb.kernel.org/~akpm/stuff/added-to-mm.txt to find out what to do about this The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/ ------------------------------------------------------ Subject: fs: use new truncate helpers From: Nick Piggin Update some fs code to make use of new helper functions introduced in the previous patch. Should be no significant change in behaviour (except CIFS now calls send_sig under i_lock, via inode_newsize_ok). Signed-off-by: Nick Piggin Reviewed-by: Christoph Hellwig Acked-by: Miklos Szeredi Cc: Cc: Trond Myklebust Cc: Cc: Steven French Cc: Jan Kara Cc: Dave Kleikamp Cc: Chris Mason Cc: OGAWA Hirofumi Cc: Hugh Dickins Cc: Steven French Cc: Al Viro Signed-off-by: Andrew Morton --- fs/buffer.c | 10 +------ fs/cifs/inode.c | 53 +++++++--------------------------------- fs/fuse/dir.c | 14 +++------- fs/fuse/fuse_i.h | 2 - fs/fuse/inode.c | 11 -------- fs/nfs/inode.c | 52 +++++++++------------------------------ fs/ramfs/file-nommu.c | 18 +++---------- 7 files changed, 34 insertions(+), 126 deletions(-) diff -puN fs/buffer.c~fs-use-new-truncate-helpers fs/buffer.c --- a/fs/buffer.c~fs-use-new-truncate-helpers +++ a/fs/buffer.c @@ -2225,16 +2225,10 @@ int generic_cont_expand_simple(struct in struct address_space *mapping = inode->i_mapping; struct page *page; void *fsdata; - unsigned long limit; int err; - err = -EFBIG; - limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur; - if (limit != RLIM_INFINITY && size > (loff_t)limit) { - send_sig(SIGXFSZ, current, 0); - goto out; - } - if (size > inode->i_sb->s_maxbytes) + err = inode_newsize_ok(inode, size); + if (err) goto out; err = pagecache_write_begin(NULL, mapping, size, 0, diff -puN fs/cifs/inode.c~fs-use-new-truncate-helpers fs/cifs/inode.c --- a/fs/cifs/inode.c~fs-use-new-truncate-helpers +++ a/fs/cifs/inode.c @@ -1557,57 +1557,24 @@ static int cifs_truncate_page(struct add static int cifs_vmtruncate(struct inode *inode, loff_t offset) { - struct address_space *mapping = inode->i_mapping; - unsigned long limit; + loff_t oldsize; + int err; spin_lock(&inode->i_lock); - if (inode->i_size < offset) - goto do_expand; - /* - * truncation of in-use swapfiles is disallowed - it would cause - * subsequent swapout to scribble on the now-freed blocks. - */ - if (IS_SWAPFILE(inode)) { + err = inode_newsize_ok(inode, offset); + if (err) { spin_unlock(&inode->i_lock); - goto out_busy; - } - i_size_write(inode, offset); - spin_unlock(&inode->i_lock); - /* - * unmap_mapping_range is called twice, first simply for efficiency - * so that truncate_inode_pages does fewer single-page unmaps. However - * after this first call, and before truncate_inode_pages finishes, - * it is possible for private pages to be COWed, which remain after - * truncate_inode_pages finishes, hence the second unmap_mapping_range - * call must be made for correctness. - */ - unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1); - truncate_inode_pages(mapping, offset); - unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1); - goto out_truncate; - -do_expand: - limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur; - if (limit != RLIM_INFINITY && offset > limit) { - spin_unlock(&inode->i_lock); - goto out_sig; - } - if (offset > inode->i_sb->s_maxbytes) { - spin_unlock(&inode->i_lock); - goto out_big; + goto out; } + + oldsize = inode->i_size; i_size_write(inode, offset); spin_unlock(&inode->i_lock); -out_truncate: + truncate_pagecache(inode, oldsize, offset); if (inode->i_op->truncate) inode->i_op->truncate(inode); - return 0; -out_sig: - send_sig(SIGXFSZ, current, 0); -out_big: - return -EFBIG; -out_busy: - return -ETXTBSY; +out: + return err; } static int diff -puN fs/fuse/dir.c~fs-use-new-truncate-helpers fs/fuse/dir.c --- a/fs/fuse/dir.c~fs-use-new-truncate-helpers +++ a/fs/fuse/dir.c @@ -1276,14 +1276,9 @@ static int fuse_do_setattr(struct dentry return 0; if (attr->ia_valid & ATTR_SIZE) { - unsigned long limit; - if (IS_SWAPFILE(inode)) - return -ETXTBSY; - limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur; - if (limit != RLIM_INFINITY && attr->ia_size > (loff_t) limit) { - send_sig(SIGXFSZ, current, 0); - return -EFBIG; - } + err = inode_newsize_ok(inode, attr->ia_size); + if (err) + return err; is_truncate = true; } @@ -1350,8 +1345,7 @@ static int fuse_do_setattr(struct dentry * FUSE_NOWRITE, otherwise fuse_launder_page() would deadlock. */ if (S_ISREG(inode->i_mode) && oldsize != outarg.attr.size) { - if (outarg.attr.size < oldsize) - fuse_truncate(inode->i_mapping, outarg.attr.size); + truncate_pagecache(inode, oldsize, outarg.attr.size); invalidate_inode_pages2(inode->i_mapping); } diff -puN fs/fuse/fuse_i.h~fs-use-new-truncate-helpers fs/fuse/fuse_i.h --- a/fs/fuse/fuse_i.h~fs-use-new-truncate-helpers +++ a/fs/fuse/fuse_i.h @@ -602,8 +602,6 @@ void fuse_change_attributes(struct inode void fuse_change_attributes_common(struct inode *inode, struct fuse_attr *attr, u64 attr_valid); -void fuse_truncate(struct address_space *mapping, loff_t offset); - /** * Initialize the client device */ diff -puN fs/fuse/inode.c~fs-use-new-truncate-helpers fs/fuse/inode.c --- a/fs/fuse/inode.c~fs-use-new-truncate-helpers +++ a/fs/fuse/inode.c @@ -121,14 +121,6 @@ static int fuse_remount_fs(struct super_ return 0; } -void fuse_truncate(struct address_space *mapping, loff_t offset) -{ - /* See vmtruncate() */ - unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1); - truncate_inode_pages(mapping, offset); - unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1); -} - void fuse_change_attributes_common(struct inode *inode, struct fuse_attr *attr, u64 attr_valid) { @@ -186,8 +178,7 @@ void fuse_change_attributes(struct inode spin_unlock(&fc->lock); if (S_ISREG(inode->i_mode) && oldsize != attr->size) { - if (attr->size < oldsize) - fuse_truncate(inode->i_mapping, attr->size); + truncate_pagecache(inode, oldsize, attr->size); invalidate_inode_pages2(inode->i_mapping); } } diff -puN fs/nfs/inode.c~fs-use-new-truncate-helpers fs/nfs/inode.c --- a/fs/nfs/inode.c~fs-use-new-truncate-helpers +++ a/fs/nfs/inode.c @@ -458,49 +458,21 @@ nfs_setattr(struct dentry *dentry, struc */ static int nfs_vmtruncate(struct inode * inode, loff_t offset) { - if (i_size_read(inode) < offset) { - unsigned long limit; + loff_t oldsize; + int err; - limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur; - if (limit != RLIM_INFINITY && offset > limit) - goto out_sig; - if (offset > inode->i_sb->s_maxbytes) - goto out_big; - spin_lock(&inode->i_lock); - i_size_write(inode, offset); - spin_unlock(&inode->i_lock); - } else { - struct address_space *mapping = inode->i_mapping; + err = inode_newsize_ok(inode, offset); + if (err) + goto out; - /* - * truncation of in-use swapfiles is disallowed - it would - * cause subsequent swapout to scribble on the now-freed - * blocks. - */ - if (IS_SWAPFILE(inode)) - return -ETXTBSY; - spin_lock(&inode->i_lock); - i_size_write(inode, offset); - spin_unlock(&inode->i_lock); + spin_lock(&inode->i_lock); + oldsize = inode->i_size; + i_size_write(inode, offset); + spin_unlock(&inode->i_lock); - /* - * unmap_mapping_range is called twice, first simply for - * efficiency so that truncate_inode_pages does fewer - * single-page unmaps. However after this first call, and - * before truncate_inode_pages finishes, it is possible for - * private pages to be COWed, which remain after - * truncate_inode_pages finishes, hence the second - * unmap_mapping_range call must be made for correctness. - */ - unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1); - truncate_inode_pages(mapping, offset); - unmap_mapping_range(mapping, offset + PAGE_SIZE - 1, 0, 1); - } - return 0; -out_sig: - send_sig(SIGXFSZ, current, 0); -out_big: - return -EFBIG; + truncate_pagecache(inode, oldsize, offset); +out: + return err; } /** diff -puN fs/ramfs/file-nommu.c~fs-use-new-truncate-helpers fs/ramfs/file-nommu.c --- a/fs/ramfs/file-nommu.c~fs-use-new-truncate-helpers +++ a/fs/ramfs/file-nommu.c @@ -69,14 +69,11 @@ int ramfs_nommu_expand_for_mapping(struc /* make various checks */ order = get_order(newsize); if (unlikely(order >= MAX_ORDER)) - goto too_big; + return -EFBIG; - limit = current->signal->rlim[RLIMIT_FSIZE].rlim_cur; - if (limit != RLIM_INFINITY && newsize > limit) - goto fsize_exceeded; - - if (newsize > inode->i_sb->s_maxbytes) - goto too_big; + ret = inode_newsize_ok(inode, newsize); + if (ret) + return ret; i_size_write(inode, newsize); @@ -118,12 +115,7 @@ int ramfs_nommu_expand_for_mapping(struc return 0; - fsize_exceeded: - send_sig(SIGXFSZ, current, 0); - too_big: - return -EFBIG; - - add_error: +add_error: while (loop < npages) __free_page(pages + loop++); return ret;