From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: [PATCH 3/4] BTRFS: Set NOSEC early for btrfs Date: Fri, 27 May 2011 15:54:04 -0700 Message-ID: <1306536845-24162-3-git-send-email-andi@firstfloor.org> References: <1306536845-24162-1-git-send-email-andi@firstfloor.org> Cc: linux-fsdevel@vger.kernel.org, viro@zeniv.linux.org.uk, Andi Kleen , josef@redhat.com, chris.mason@oracle.com To: linux-kernel@vger.kernel.org Return-path: In-Reply-To: <1306536845-24162-1-git-send-email-andi@firstfloor.org> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org From: Andi Kleen Based on a suggestion from Josef Bacik. The btrfs inode lookup already does a quick check for xattrs. If it doesn't find any and the file is not sgid set the NOSEC flag. This avoids a xattr lookup on the first write. Cc: josef@redhat.com Cc: chris.mason@oracle.com Signed-off-by: Andi Kleen --- fs/btrfs/inode.c | 7 ++++++- 1 files changed, 6 insertions(+), 1 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 7cd8ab0..b309c30 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2544,8 +2544,11 @@ static void btrfs_read_locked_inode(struct inode *inode) * any xattrs or acls */ maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino); - if (!maybe_acls) + if (!maybe_acls) { cache_no_acl(inode); + if (!is_sgid(inode->i_mode)) + inode->i_flags |= S_NOSEC; + } BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0, alloc_group_block, 0); @@ -4616,6 +4619,8 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, (BTRFS_I(dir)->flags & BTRFS_INODE_NODATACOW)) BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; } + if (!is_sgid(mode)) + inode->i_flags |= S_NOSEC; insert_inode_hash(inode); inode_tree_add(inode); -- 1.7.4.4