From 0aefb1183fa121180025bf4fa83ce81825f2ecfa Mon Sep 17 00:00:00 2001 From: Ben Collins <ben.collins@canonical.com> Date: Sun, 27 Jul 2008 22:30:16 -0400 Subject: [PATCH] UBUNTU: ubuntu: unionfs: Added v1.4 module from hardy Took some porting... [ adapt for mandriva kernel -- herton ] Signed-off-by: Ben Collins <ben.collins@canonical.com> Signed-off-by: Herton Ronaldo Krzesinski <herton@mandriva.com.br> --- fs/Kconfig | 1 fs/Makefile | 1 fs/unionfs/Kconfig | 8 fs/unionfs/Makefile | 16 fs/unionfs/branchman.c | 579 +++++++++++++++++++++++ fs/unionfs/commonfops.c | 706 ++++++++++++++++++++++++++++ fs/unionfs/copyup.c | 738 +++++++++++++++++++++++++++++ fs/unionfs/dentry.c | 282 +++++++++++ fs/unionfs/dirfops.c | 330 +++++++++++++ fs/unionfs/dirhelper.c | 309 ++++++++++++ fs/unionfs/file.c | 393 +++++++++++++++ fs/unionfs/inode.c | 1034 ++++++++++++++++++++++++++++++++++++++++++ fs/unionfs/lookup.c | 524 +++++++++++++++++++++ fs/unionfs/main.c | 869 +++++++++++++++++++++++++++++++++++ fs/unionfs/persistent_inode.c | 658 ++++++++++++++++++++++++++ fs/unionfs/print.c | 439 +++++++++++++++++ fs/unionfs/rdstate.c | 328 +++++++++++++ fs/unionfs/rename.c | 943 ++++++++++++++++++++++++++++++++++++++ fs/unionfs/sioq.c | 109 ++++ fs/unionfs/sioq.h | 80 +++ fs/unionfs/stale_inode.c | 133 +++++ fs/unionfs/subr.c | 204 ++++++++ fs/unionfs/super.c | 762 ++++++++++++++++++++++++++++++ fs/unionfs/unionfs.h | 736 +++++++++++++++++++++++++++++ fs/unionfs/unionfs_debug.h | 191 +++++++ fs/unionfs/unionfs_imap.h | 92 +++ fs/unionfs/unionfs_macros.h | 206 ++++++++ fs/unionfs/unlink.c | 380 +++++++++++++++ fs/unionfs/xattr.c | 165 ++++++ 29 files changed, 11216 insertions(+) --- a/fs/Kconfig +++ b/fs/Kconfig @@ -282,6 +282,7 @@ source "fs/partitions/Kconfig" endmenu endif +source "fs/unionfs/Kconfig" source "fs/nls/Kconfig" source "fs/dlm/Kconfig" --- a/fs/Makefile +++ b/fs/Makefile @@ -119,6 +119,7 @@ obj-$(CONFIG_9P_FS) += 9p/ obj-$(CONFIG_AFS_FS) += afs/ obj-$(CONFIG_BEFS_FS) += befs/ obj-$(CONFIG_HOSTFS) += hostfs/ +obj-$(CONFIG_UNION_FS) += unionfs/ obj-$(CONFIG_HPPFS) += hppfs/ obj-$(CONFIG_DEBUG_FS) += debugfs/ obj-$(CONFIG_OCFS2_FS) += ocfs2/ --- /dev/null +++ b/fs/unionfs/Kconfig @@ -0,0 +1,8 @@ +config UNION_FS + tristate "UnionFS 1.4 Legacy" + default m + help + This is the v1.4 module, forward ported and beaten into + AppArmor submission. We know this one works for the LiveCD, so + this is to keep the release managers from having to spend + long nights and pushing off deadlines. --- /dev/null +++ b/fs/unionfs/Makefile @@ -0,0 +1,16 @@ +UNIONFS_VERSION = 1.4 +SUP_MAJOR= 2 +SUP_MINOR= 6 +SUP_PATCH= 18 + +EXTRA_CFLAGS+=-DUNIONFS_VERSION=\"${UNIONFS_VERSION}\" -DSUP_MAJOR=${SUP_MAJOR} -DSUP_MINOR=${SUP_MINOR} -DSUP_PATCH=${SUP_PATCH} -DUNIONFS_UNSUPPORTED + +# This will enable full debugging support +# EXTRA_CFLAGS+=-DUNIONFS_DEBUG + +obj-$(CONFIG_UNION_FS) += unionfs.o + +unionfs-objs := subr.o dentry.o file.o inode.o main.o super.o \ + stale_inode.o branchman.o xattr.o rdstate.o copyup.o \ + dirhelper.o rename.o unlink.o lookup.o persistent_inode.o \ + commonfops.o dirfops.o print.o sioq.o --- /dev/null +++ b/fs/unionfs/branchman.c @@ -0,0 +1,579 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: branchman.c,v 1.66 2006/10/31 00:05:22 yiannos Exp $ + */ + +#include "unionfs.h" + +struct dentry **alloc_new_dentries(int objs) +{ + if (!objs) + return NULL; + + return KZALLOC(sizeof(struct dentry *) * objs, GFP_KERNEL); +} + +struct unionfs_usi_data *alloc_new_data(int objs) +{ + if (!objs) + return NULL; + + return KZALLOC(sizeof(struct unionfs_usi_data) * objs, GFP_KERNEL); +} + +static void fixputmaps(struct super_block *sb) +{ + struct unionfs_sb_info *spd; + struct putmap *cur; + int gen; + int i; + + print_entry_location(); + + spd = stopd(sb); + cur = spd->usi_putmaps[spd->usi_lastputmap - spd->usi_firstputmap]; + + for (gen = 0; gen < spd->usi_lastputmap - spd->usi_firstputmap; gen++) { + if (!spd->usi_putmaps[gen]) + continue; + for (i = 0; i <= spd->usi_putmaps[gen]->bend; i++) + spd->usi_putmaps[gen]->map[i] = + cur->map[spd->usi_putmaps[gen]->map[i]]; + } + + print_exit_location(); +} + +static int newputmap(struct super_block *sb) +{ + struct unionfs_sb_info *spd; + struct putmap *newmap; + int count = 0; + int i; + + print_entry_location(); + + spd = stopd(sb); + + i = sizeof(int) * (sbend(sb) + 1); + newmap = KMALLOC(sizeof(struct putmap) + i, GFP_KERNEL); + if (!newmap) { + print_exit_status(-ENOMEM); + return -ENOMEM; + } + + if (!spd->usi_firstputmap) { + spd->usi_firstputmap = 1; + spd->usi_lastputmap = 1; + + spd->usi_putmaps = KMALLOC(sizeof(struct putmap *), GFP_KERNEL); + if (!spd->usi_putmaps) { + KFREE(newmap); + print_exit_status(-ENOMEM); + return -ENOMEM; + } + } else { + struct putmap **newlist; + int newfirst = spd->usi_firstputmap; + + while (!spd->usi_putmaps[newfirst - spd->usi_firstputmap] && + newfirst <= spd->usi_lastputmap) { + newfirst++; + } + + newlist = + KMALLOC(sizeof(struct putmap *) * + (1 + spd->usi_lastputmap - newfirst), GFP_KERNEL); + if (!newlist) { + KFREE(newmap); + print_exit_status(-ENOMEM); + return -ENOMEM; + } + + for (i = newfirst; i <= spd->usi_lastputmap; i++) { + newlist[i - newfirst] = + spd->usi_putmaps[i - spd->usi_firstputmap]; + } + + KFREE(spd->usi_putmaps); + spd->usi_putmaps = newlist; + spd->usi_firstputmap = newfirst; + spd->usi_lastputmap++; + } + + newmap->bend = sbend(sb); + for (i = 0; i <= sbend(sb); i++) { + count += branch_count(sb, i); + newmap->map[i] = i; + } + for (i = spd->usi_firstputmap; i < spd->usi_lastputmap; i++) { + struct putmap *cur; + cur = spd->usi_putmaps[i - spd->usi_firstputmap]; + if (!cur) + continue; + count -= atomic_read(&cur->count); + } + atomic_set(&newmap->count, count); + spd->usi_putmaps[spd->usi_lastputmap - spd->usi_firstputmap] = newmap; + + print_exit_status(0); + return 0; +} + +/* XXX: this function needs to go. There is no reason for this to be here */ +int unionfs_ioctl_branchcount(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int err = 0; + int bstart, bend; + int i; + struct super_block *sb = file->f_dentry->d_sb; + + print_entry_location(); + + bstart = sbstart(sb); + bend = sbend(sb); + + err = bend + 1; + if (!arg) + goto out; + + for (i = bstart; i <= bend; i++) { + if (put_user(branch_count(sb, i), ((int __user *)arg) + i)) { + err = -EFAULT; + goto out; + } + } + + out: + print_exit_status(err); + return err; +} + +int unionfs_ioctl_incgen(struct file *file, unsigned int cmd, unsigned long arg) +{ + int err = 0; + struct super_block *sb; + + print_entry_location(); + + sb = file->f_dentry->d_sb; + + unionfs_write_lock(sb); + if ((err = newputmap(sb))) + goto out; + + atomic_inc(&stopd(sb)->usi_generation); + err = atomic_read(&stopd(sb)->usi_generation); + + atomic_set(&dtopd(sb->s_root)->udi_generation, err); + atomic_set(&itopd(sb->s_root->d_inode)->uii_generation, err); + + out: + unionfs_write_unlock(sb); + print_exit_status(err); + return err; +} + +int unionfs_ioctl_addbranch(struct inode *inode, unsigned int cmd, + unsigned long arg) +{ + int err; + struct unionfs_addbranch_args *addargs = NULL; + struct nameidata nd; + char *path = NULL; + int gen; + int i; + + int pobjects; + + struct unionfs_usi_data *new_data = NULL; + struct dentry **new_udi_dentry = NULL; + struct inode **new_uii_inode = NULL; + + struct dentry *root = NULL; + struct dentry *hidden_root = NULL; + + print_entry_location(); + +#ifdef UNIONFS_IMAP + if (stopd(sb)->usi_persistent) { + printk(KERN_ERR "Cannot manipulate branches if imap is used\n"); + err = -EPERM; + goto out; + } +#endif + + err = -ENOMEM; + addargs = KMALLOC(sizeof(struct unionfs_addbranch_args), GFP_KERNEL); + if (!addargs) + goto out; + + err = -EFAULT; + if (copy_from_user + (addargs, (const void __user *)arg, + sizeof(struct unionfs_addbranch_args))) + goto out; + + err = -EINVAL; + if (addargs->ab_perms & ~(MAY_READ | MAY_WRITE | MAY_NFSRO)) + goto out; + if (!(addargs->ab_perms & MAY_READ)) + goto out; + + err = -E2BIG; + if (sbend(inode->i_sb) > FD_SETSIZE) + goto out; + + err = -ENOMEM; + if (!(path = getname((const char __user *)addargs->ab_path))) + goto out; + + err = path_lookup(path, LOOKUP_FOLLOW, &nd); + + RECORD_PATH_LOOKUP(&nd); + if (err) + goto out; + if ((err = check_branch(&nd))) { + path_put(&nd.path); + RECORD_PATH_RELEASE(&nd); + goto out; + } + + unionfs_write_lock(inode->i_sb); + lock_dentry(inode->i_sb->s_root); + + root = inode->i_sb->s_root; + for (i = dbstart(inode->i_sb->s_root); i <= dbend(inode->i_sb->s_root); + i++) { + hidden_root = dtohd_index(root, i); + if (is_branch_overlap(hidden_root, nd.path.dentry)) { + err = -EINVAL; + goto out; + } + } + + err = -EINVAL; + if (addargs->ab_branch < 0 + || (addargs->ab_branch > (sbend(inode->i_sb) + 1))) + goto out; + + if ((err = newputmap(inode->i_sb))) + goto out; + + stopd(inode->i_sb)->b_end++; + dtopd(inode->i_sb->s_root)->udi_bcount++; + set_dbend(inode->i_sb->s_root, dbend(inode->i_sb->s_root) + 1); + itopd(inode->i_sb->s_root->d_inode)->b_end++; + + atomic_inc(&stopd(inode->i_sb)->usi_generation); + gen = atomic_read(&stopd(inode->i_sb)->usi_generation); + + pobjects = sbend(inode->i_sb) + 1; + + /* Reallocate the dynamic structures. */ + new_data = alloc_new_data(pobjects); + new_udi_dentry = alloc_new_dentries(pobjects); + new_uii_inode = KZALLOC(sizeof(struct inode *) * pobjects, GFP_KERNEL); + + if (!new_udi_dentry || !new_uii_inode || !new_data) { + err = -ENOMEM; + goto out; + } + + /* Copy the in-place values to our new structure. */ + for (i = 0; i < addargs->ab_branch; i++) { + atomic_set(&(new_data[i].sbcount), + branch_count(inode->i_sb, i)); + + new_data[i].branchperms = branchperms(inode->i_sb, i); + new_data[i].hidden_mnt = stohiddenmnt_index(inode->i_sb, i); + new_data[i].sb = stohs_index(inode->i_sb, i); + + new_udi_dentry[i] = dtohd_index(inode->i_sb->s_root, i); + new_uii_inode[i] = itohi_index(inode->i_sb->s_root->d_inode, i); + } + + /* Shift the ends to the right (only handle reallocated bits). */ + for (i = sbend(inode->i_sb) - 1; i >= (int)addargs->ab_branch; i--) { + int j = i + 1; + int pmindex; + + atomic_set(&new_data[j].sbcount, branch_count(inode->i_sb, i)); + + new_data[j].branchperms = branchperms(inode->i_sb, i); + new_data[j].hidden_mnt = stohiddenmnt_index(inode->i_sb, i); + new_data[j].sb = stohs_index(inode->i_sb, i); + new_udi_dentry[j] = dtohd_index(inode->i_sb->s_root, i); + new_uii_inode[j] = itohi_index(inode->i_sb->s_root->d_inode, i); + + /* Update the newest putmap, so it is correct for later. */ + pmindex = stopd(inode->i_sb)->usi_lastputmap; + pmindex -= stopd(inode->i_sb)->usi_firstputmap; + stopd(inode->i_sb)->usi_putmaps[pmindex]->map[i] = j; + + } + + /* Now we can free the old ones. */ + KFREE(dtopd(inode->i_sb->s_root)->udi_dentry); + KFREE(itopd(inode->i_sb->s_root->d_inode)->uii_inode); + KFREE(stopd(inode->i_sb)->usi_data); + + /* Update the real pointers. */ + dtohd_ptr(inode->i_sb->s_root) = new_udi_dentry; + itohi_ptr(inode->i_sb->s_root->d_inode) = new_uii_inode; + stopd(inode->i_sb)->usi_data = new_data; + + /* Re-NULL the new ones so we don't try to free them. */ + new_data = NULL; + new_udi_dentry = NULL; + new_uii_inode = NULL; + + /* Put the new dentry information into it's slot. */ + set_dtohd_index(inode->i_sb->s_root, addargs->ab_branch, nd.path.dentry); + set_itohi_index(inode->i_sb->s_root->d_inode, addargs->ab_branch, + IGRAB(nd.path.dentry->d_inode)); + set_branchperms(inode->i_sb, addargs->ab_branch, addargs->ab_perms); + set_branch_count(inode->i_sb, addargs->ab_branch, 0); + set_stohiddenmnt_index(inode->i_sb, addargs->ab_branch, nd.path.mnt); + set_stohs_index(inode->i_sb, addargs->ab_branch, nd.path.dentry->d_sb); + + atomic_set(&dtopd(inode->i_sb->s_root)->udi_generation, gen); + atomic_set(&itopd(inode->i_sb->s_root->d_inode)->uii_generation, gen); + + fixputmaps(inode->i_sb); + + out: + unlock_dentry(inode->i_sb->s_root); + unionfs_write_unlock(inode->i_sb); + + KFREE(new_udi_dentry); + KFREE(new_uii_inode); + KFREE(new_data); + KFREE(addargs); + if (path) + putname(path); + + print_exit_status(err); + + return err; +} + +/* This must be called with the super block already locked. */ +int unionfs_ioctl_delbranch(struct super_block *sb, unsigned long arg) +{ + struct dentry *hidden_dentry; + struct inode *hidden_inode; + struct vfsmount *hidden_mnt; + struct dentry *root_dentry; + struct inode *root_inode; + int err = 0; + int pmindex, i, gen; + + print_entry("branch = %lu ", arg); + lock_dentry(sb->s_root); + +#ifdef UNIONFS_IMAP + if (stopd(sb)->usi_persistent) { + printk(KERN_ERR "Cannot manipulate branches if imap is used\n"); + err = -EPERM; + goto out; + } +#endif + err = -EBUSY; + if (sbmax(sb) == 1) + goto out; + err = -EINVAL; + if (arg < 0 || arg > stopd(sb)->b_end) + goto out; + err = -EBUSY; + if (branch_count(sb, arg)) + goto out; + if ((err = newputmap(sb))) + goto out; + + pmindex = stopd(sb)->usi_lastputmap; + pmindex -= stopd(sb)->usi_firstputmap; + + atomic_inc(&stopd(sb)->usi_generation); + gen = atomic_read(&stopd(sb)->usi_generation); + + root_dentry = sb->s_root; + root_inode = sb->s_root->d_inode; + + hidden_dentry = dtohd_index(root_dentry, arg); + hidden_mnt = stohiddenmnt_index(sb, arg); + hidden_inode = itohi_index(root_inode, arg); + + DPUT(hidden_dentry); + IPUT(hidden_inode); + mntput(hidden_mnt); + + for (i = arg; i <= (sbend(sb) - 1); i++) { + set_branch_count(sb, i, branch_count(sb, i + 1)); + set_stohiddenmnt_index(sb, i, stohiddenmnt_index(sb, i + 1)); + set_stohs_index(sb, i, stohs_index(sb, i + 1)); + set_branchperms(sb, i, branchperms(sb, i + 1)); + set_dtohd_index(root_dentry, i, + dtohd_index(root_dentry, i + 1)); + set_itohi_index(root_inode, i, itohi_index(root_inode, i + 1)); + stopd(sb)->usi_putmaps[pmindex]->map[i + 1] = i; + } + + set_dtohd_index(root_dentry, sbend(sb), NULL); + set_itohi_index(root_inode, sbend(sb), NULL); + set_stohiddenmnt_index(sb, sbend(sb), NULL); + set_stohs_index(sb, sbend(sb), NULL); + + //XXX: Place check for inode maps and removal of branch here + + stopd(sb)->b_end--; + set_dbend(root_dentry, dbend(root_dentry) - 1); + dtopd(root_dentry)->udi_bcount--; + itopd(root_inode)->b_end--; + + atomic_set(&dtopd(root_dentry)->udi_generation, gen); + atomic_set(&itopd(root_inode)->uii_generation, gen); + + fixputmaps(sb); + + /* This doesn't open a file, so we might have to free the map here. */ + if (atomic_read(&stopd(sb)->usi_putmaps[pmindex]->count) == 0) { + KFREE(stopd(sb)->usi_putmaps[pmindex]); + stopd(sb)->usi_putmaps[pmindex] = NULL; + } + + out: + unlock_dentry(sb->s_root); + print_exit_status(err); + + return err; +} + +int unionfs_ioctl_rdwrbranch(struct inode *inode, unsigned int cmd, + unsigned long arg) +{ + int err; + struct unionfs_rdwrbranch_args *rdwrargs = NULL; + int gen; + + print_entry_location(); + + unionfs_write_lock(inode->i_sb); + lock_dentry(inode->i_sb->s_root); + + if ((err = newputmap(inode->i_sb))) + goto out; + + err = -ENOMEM; + rdwrargs = KMALLOC(sizeof(struct unionfs_rdwrbranch_args), GFP_KERNEL); + if (!rdwrargs) + goto out; + + err = -EFAULT; + if (copy_from_user + (rdwrargs, (const void __user *)arg, + sizeof(struct unionfs_rdwrbranch_args))) + goto out; + + err = -EINVAL; + if (rdwrargs->rwb_branch < 0 + || (rdwrargs->rwb_branch > (sbend(inode->i_sb) + 1))) + goto out; + if (rdwrargs->rwb_perms & ~(MAY_READ | MAY_WRITE | MAY_NFSRO)) + goto out; + if (!(rdwrargs->rwb_perms & MAY_READ)) + goto out; + + set_branchperms(inode->i_sb, rdwrargs->rwb_branch, rdwrargs->rwb_perms); + + atomic_inc(&stopd(inode->i_sb)->usi_generation); + gen = atomic_read(&stopd(inode->i_sb)->usi_generation); + atomic_set(&dtopd(inode->i_sb->s_root)->udi_generation, gen); + atomic_set(&itopd(inode->i_sb->s_root->d_inode)->uii_generation, gen); + + err = 0; + + out: + unlock_dentry(inode->i_sb->s_root); + unionfs_write_unlock(inode->i_sb); + KFREE(rdwrargs); + + print_exit_status(err); + + return err; +} + +int unionfs_ioctl_queryfile(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int err = 0; + fd_set branchlist; + + int bstart = 0, bend = 0, bindex = 0; + struct dentry *dentry, *hidden_dentry; + + print_entry_location(); + + dentry = file->f_dentry; + lock_dentry(dentry); + if ((err = unionfs_partial_lookup(dentry))) + goto out; + bstart = dbstart(dentry); + bend = dbend(dentry); + + FD_ZERO(&branchlist); + + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + if (hidden_dentry->d_inode) + FD_SET(bindex, &branchlist); + } + + err = copy_to_user((void __user *)arg, &branchlist, sizeof(fd_set)); + if (err) { + err = -EFAULT; + goto out; + } + + out: + unlock_dentry(dentry); + err = err < 0 ? err : bend; + print_exit_status(err); + return (err); +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/commonfops.c @@ -0,0 +1,706 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: commonfops.c,v 1.61 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +/* We only need this function here, but it could get promoted to unionfs.h, if + * other things need a generation specific branch putting function. */ +static inline void branchput_gen(int generation, struct super_block *sb, + int index) +{ + struct putmap *putmap; + + if (generation == atomic_read(&stopd(sb)->usi_generation)) { + branchput(sb, index); + return; + } + + BUG_ON(stopd(sb)->usi_firstputmap > generation); + BUG_ON(stopd(sb)->usi_lastputmap < generation); + + putmap = + stopd(sb)->usi_putmaps[generation - stopd(sb)->usi_firstputmap]; + BUG_ON(index < 0); + BUG_ON(index > putmap->bend); + BUG_ON(putmap->map[index] < 0); + branchput(sb, putmap->map[index]); + if (atomic_dec_and_test(&putmap->count)) { + stopd(sb)->usi_putmaps[generation - stopd(sb)->usi_firstputmap] + = NULL; + dprint(PRINT_DEBUG, "Freeing putmap %d.\n", generation); + KFREE(putmap); + } +} + +static char *get_random_name(int size, unsigned char *name) +{ + int i; + int j; + unsigned char *tmpbuf = NULL; + + if (size <= WHLEN) + return NULL; + + if (!name) + name = KMALLOC(size + 1, GFP_KERNEL); + if (!name) { + name = ERR_PTR(-ENOMEM); + goto out; + } + strncpy(name, WHPFX, WHLEN); + + tmpbuf = KMALLOC(size, GFP_KERNEL); + if (!tmpbuf) { + KFREE(name); + name = ERR_PTR(-ENOMEM); + goto out; + } + + get_random_bytes((void *)tmpbuf, (size - 3) / 2); + + j = WHLEN; + i = 0; + while ((i < (size - 3) / 2) && (j < size)) { + /* get characters in the 0-9, A-F range */ + + name[j] = + (tmpbuf[i] % 16) < + 10 ? (tmpbuf[i] % 16) + '0' : (tmpbuf[i] % 16) + 'a'; + j++; + if (j == size) + break; + name[j] = + (tmpbuf[i] >> 4) < + 10 ? (tmpbuf[i] >> 4) + '0' : (tmpbuf[i] >> 4) + 'a'; + j++; + + i++; + } + + name[size] = '\0'; + + out: + KFREE(tmpbuf); + return (name); + +} + +static int copyup_deleted_file(struct file *file, struct dentry *dentry, + int bstart, int bindex) +{ + int attempts = 0; + int err; + int exists = 1; + char *name = NULL; + struct dentry *tmp_dentry = NULL; + struct dentry *hidden_dentry = NULL; + struct dentry *hidden_dir_dentry = NULL; + + print_entry_location(); + + /* Try five times to get a unique file name, fail after that. Five is + * simply a magic number, because we shouldn't try forever. */ + while (exists) { + /* The first call allocates, the subsequent ones reuse. */ + name = get_random_name(UNIONFS_TMPNAM_LEN, name); + err = -ENOMEM; + if (!name) + goto out; + //XXX: Why do we do this every time? bstart never changes? + hidden_dentry = dtohd_index(dentry, bstart); + + tmp_dentry = LOOKUP_ONE_LEN(name, hidden_dentry->d_parent, + UNIONFS_TMPNAM_LEN); + err = PTR_ERR(tmp_dentry); + if (IS_ERR(tmp_dentry)) + goto out; + exists = tmp_dentry->d_inode ? 1 : 0; + DPUT(tmp_dentry); + + err = -EEXIST; + if (++attempts > 5) + goto out; + } + + err = copyup_named_file(dentry->d_parent->d_inode, file, name, bstart, + bindex, file->f_dentry->d_inode->i_size); + if (err) + goto out; + + /* bring it to the same state as an unlinked file */ + hidden_dentry = dtohd_index(dentry, dbstart(dentry)); + hidden_dir_dentry = lock_parent(hidden_dentry); + err = vfs_unlink(hidden_dir_dentry->d_inode, hidden_dentry, NULL); + unlock_dir(hidden_dir_dentry); + + out: + KFREE(name); + print_exit_status(err); + return err; +} + +int unionfs_file_revalidate(struct file *file, int willwrite) +{ + struct super_block *sb; + struct dentry *dentry; + int sbgen, fgen, dgen; + int bindex, bstart, bend; + struct file *hidden_file; + struct dentry *hidden_dentry; + int size; + + int err = 0; + + print_entry(" file = %p", file); + + dentry = file->f_dentry; + lock_dentry(dentry); + sb = dentry->d_sb; + unionfs_read_lock(sb); + if (!unionfs_d_revalidate(dentry, NULL) && !d_deleted(dentry)) { + err = -ESTALE; + goto out; + } + print_dentry("file revalidate in", dentry); + + sbgen = atomic_read(&stopd(sb)->usi_generation); + dgen = atomic_read(&dtopd(dentry)->udi_generation); + fgen = atomic_read(&ftopd(file)->ufi_generation); + + BUG_ON(sbgen > dgen); + + /* There are two cases we are interested in. The first is if the + * generation is lower than the super-block. The second is if someone + * has copied up this file from underneath us, we also need to refresh + * things. */ + if (!d_deleted(dentry) && + ((sbgen > fgen) || (dbstart(dentry) != fbstart(file)))) { + /* First we throw out the existing files. */ + bstart = fbstart(file); + bend = fbend(file); + for (bindex = bstart; bindex <= bend; bindex++) { + if (ftohf_index(file, bindex)) { + branchput_gen(fgen, dentry->d_sb, bindex); + fput(ftohf_index(file, bindex)); + } + } + + if (ftohf_ptr(file)) { + KFREE(ftohf_ptr(file)); + ftohf_ptr(file) = NULL; + } + + /* Now we reopen the file(s) as in unionfs_open. */ + bstart = fbstart(file) = dbstart(dentry); + bend = fbend(file) = dbend(dentry); + + size = sizeof(struct file *) * sbmax(sb); + ftohf_ptr(file) = KZALLOC(size, GFP_KERNEL); + if (!ftohf_ptr(file)) { + err = -ENOMEM; + goto out; + } + + if (S_ISDIR(dentry->d_inode->i_mode)) { + /* We need to open all the files. */ + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + + DGET(hidden_dentry); + mntget(stohiddenmnt_index(sb, bindex)); + branchget(sb, bindex); + + hidden_file = + DENTRY_OPEN(hidden_dentry, + stohiddenmnt_index(sb, bindex), + file->f_flags); + if (IS_ERR(hidden_file)) { + err = PTR_ERR(hidden_file); + goto out; + } else { + set_ftohf_index(file, bindex, + hidden_file); + } + } + } else { + /* We only open the highest priority branch. */ + hidden_dentry = dtohd(dentry); + if (willwrite && IS_WRITE_FLAG(file->f_flags) + && is_robranch(dentry)) { + for (bindex = bstart - 1; bindex >= 0; bindex--) { + + err = copyup_file(dentry-> + d_parent-> + d_inode, + file, + bstart, + bindex, + file-> + f_dentry-> + d_inode->i_size); + + if (!err) + break; + else + continue; + + } + atomic_set(&ftopd(file)->ufi_generation, + atomic_read(&itopd(dentry->d_inode)-> + uii_generation)); + goto out; + } + + DGET(hidden_dentry); + mntget(stohiddenmnt_index(sb, bstart)); + branchget(sb, bstart); + hidden_file = + DENTRY_OPEN(hidden_dentry, + stohiddenmnt_index(sb, bstart), + file->f_flags); + if (IS_ERR(hidden_file)) { + err = PTR_ERR(hidden_file); + goto out; + } + set_ftohf(file, hidden_file); + /* Fix up the position. */ + hidden_file->f_pos = file->f_pos; + + memcpy(&(hidden_file->f_ra), &(file->f_ra), + sizeof(struct file_ra_state)); + } + atomic_set(&ftopd(file)->ufi_generation, + atomic_read(&itopd(dentry->d_inode)-> + uii_generation)); + } + + /* Copyup on the first write to a file on a readonly branch. */ + if (willwrite && IS_WRITE_FLAG(file->f_flags) + && !IS_WRITE_FLAG(ftohf(file)->f_flags) && is_robranch(dentry)) { + dprint(PRINT_DEBUG, + "Doing delayed copyup of a read-write file on a read-only branch.\n"); + bstart = fbstart(file); + bend = fbend(file); + + BUG_ON(!S_ISREG(file->f_dentry->d_inode->i_mode)); + + for (bindex = bstart - 1; bindex >= 0; bindex--) { + if (!d_deleted(file->f_dentry)) { + err = + copyup_file(dentry->d_parent-> + d_inode, file, bstart, + bindex, + file->f_dentry-> + d_inode->i_size); + } else { + err = + copyup_deleted_file(file, dentry, bstart, + bindex); + } + + if (!err) + break; + else + continue; + + } + if (!err && (bstart > fbstart(file))) { + bend = fbend(file); + for (bindex = bstart; bindex <= bend; bindex++) { + if (ftohf_index(file, bindex)) { + branchput(dentry->d_sb, bindex); + fput(ftohf_index(file, bindex)); + set_ftohf_index(file, bindex, NULL); + } + } + fbend(file) = bend; + } + } + + out: + print_dentry("file revalidate out", dentry); + unlock_dentry(dentry); + unionfs_read_unlock(dentry->d_sb); + print_exit_status(err); + return err; +} + +int unionfs_open(struct inode *inode, struct file *file) +{ + int err = 0; + int hidden_flags; + struct file *hidden_file = NULL; + struct dentry *hidden_dentry = NULL; + struct dentry *dentry = NULL; + int bindex = 0, bstart = 0, bend = 0; + int locked = 0; + int size; + + print_entry_location(); + + ftopd_lhs(file) = KZALLOC(sizeof(struct unionfs_file_info), GFP_KERNEL); + if (!ftopd(file)) { + err = -ENOMEM; + goto out; + } + fbstart(file) = -1; + fbend(file) = -1; + atomic_set(&ftopd(file)->ufi_generation, + atomic_read(&itopd(inode)->uii_generation)); + + size = sizeof(struct file *) * sbmax(inode->i_sb); + ftohf_ptr(file) = KZALLOC(size, GFP_KERNEL); + if (!ftohf_ptr(file)) { + err = -ENOMEM; + goto out; + } + + hidden_flags = file->f_flags; + + dentry = file->f_dentry; + dprint(PRINT_DEBUG, "dentry to open is %p\n", dentry); + lock_dentry(dentry); + unionfs_read_lock(inode->i_sb); + locked = 1; + + bstart = fbstart(file) = dbstart(dentry); + bend = fbend(file) = dbend(dentry); + + /* increment to show the kind of open, so that we can + * flush appropriately + */ + atomic_inc(&itopd(dentry->d_inode)->uii_totalopens); + + /* open all directories and make the unionfs file struct point to these hidden file structs */ + if (S_ISDIR(inode->i_mode)) { + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + + DGET(hidden_dentry); + mntget(stohiddenmnt_index(inode->i_sb, bindex)); + hidden_file = + DENTRY_OPEN(hidden_dentry, + stohiddenmnt_index(inode->i_sb, bindex), + hidden_flags); + if (IS_ERR(hidden_file)) { + err = PTR_ERR(hidden_file); + goto out; + } + + set_ftohf_index(file, bindex, hidden_file); + /* The branchget goes after the open, because otherwise + * we would miss the reference on release. */ + branchget(inode->i_sb, bindex); + } + } else { + /* open a file */ + hidden_dentry = dtohd(dentry); + + /* check for the permission for hidden file. If the error is COPYUP_ERR, + * copyup the file. + */ + if (hidden_dentry->d_inode && is_robranch(dentry)) { + /* if the open will change the file, copy it up otherwise defer it. */ + if (hidden_flags & O_TRUNC) { + int size = 0; + + err = -EROFS; + /* copyup the file */ + for (bindex = bstart - 1; bindex >= 0; bindex--) { + err = + copyup_file(dentry-> + d_parent-> + d_inode, file, + bstart, bindex, size); + if (!err) { + break; + } + } + goto out; + } else { + hidden_flags &= ~(OPEN_WRITE_FLAGS); + } + } + + DGET(hidden_dentry); + /* dentry_open will decrement mnt refcnt if err. + * otherwise fput() will do an mntput() for us upon file close. + */ + mntget(stohiddenmnt_index(inode->i_sb, bstart)); + hidden_file = DENTRY_OPEN(hidden_dentry, + stohiddenmnt_index(inode->i_sb, + bstart), + hidden_flags); + if (IS_ERR(hidden_file)) { + err = PTR_ERR(hidden_file); + goto out; + } else { + set_ftohf(file, hidden_file); + branchget(inode->i_sb, bstart); + } + } + + out: + /* freeing the allocated resources, and fput the opened files */ + if (err < 0 && ftopd(file)) { + if (!locked) + unionfs_read_lock(file->f_dentry->d_sb); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_file = ftohf_index(file, bindex); + if (hidden_file) { + branchput(file->f_dentry->d_sb, bindex); + /* fput calls dput for hidden_dentry */ + fput(hidden_file); + } + } + if (!locked) + unionfs_read_unlock(file->f_dentry->d_sb); + KFREE(ftohf_ptr(file)); + KFREE(ftopd(file)); + } + + print_file("OUT: unionfs_open", file); + + if (locked) { + unlock_dentry(dentry); + unionfs_read_unlock(inode->i_sb); + } + print_exit_status(err); + return err; +} + +int unionfs_file_release(struct inode *inode, struct file *file) +{ + int err = 0; + struct file *hidden_file = NULL; + int bindex, bstart, bend; + int fgen; + + print_entry_location(); + + checkinode(inode, "unionfs_release"); + + /* fput all the hidden files */ + fgen = atomic_read(&ftopd(file)->ufi_generation); + bstart = fbstart(file); + bend = fbend(file); + + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_file = ftohf_index(file, bindex); + + if (hidden_file) { + fput(hidden_file); + unionfs_read_lock(inode->i_sb); + branchput_gen(fgen, inode->i_sb, bindex); + unionfs_read_unlock(inode->i_sb); + } + } + KFREE(ftohf_ptr(file)); + + if (ftopd(file)->rdstate) { + ftopd(file)->rdstate->uds_access = jiffies; + dprint(PRINT_DEBUG, "Saving rdstate with cookie %u [%d.%lld]\n", + ftopd(file)->rdstate->uds_cookie, + ftopd(file)->rdstate->uds_bindex, + (long long)ftopd(file)->rdstate->uds_dirpos); + spin_lock(&itopd(inode)->uii_rdlock); + itopd(inode)->uii_rdcount++; + list_add_tail(&ftopd(file)->rdstate->uds_cache, + &itopd(inode)->uii_readdircache); + mark_inode_dirty(inode); + spin_unlock(&itopd(inode)->uii_rdlock); + ftopd(file)->rdstate = NULL; + } + KFREE(ftopd(file)); + + checkinode(inode, "post unionfs_release"); + + print_exit_status(err); + return err; +} + +long unionfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + long err = 0; /* don't fail by default */ + struct file *hidden_file = NULL; + int val; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 1))) + goto out; + + /* check if asked for local commands */ + switch (cmd) { + case FIST_IOCTL_GET_DEBUG_VALUE: + if (!capable(CAP_SYS_ADMIN)) { + err = -EACCES; + goto out; + } + val = get_debug_mask(); + err = put_user(val, (int __user *)arg); + break; + + case FIST_IOCTL_SET_DEBUG_VALUE: + if (!capable(CAP_SYS_ADMIN)) { + err = -EACCES; + goto out; + } + err = get_user(val, (int __user *)arg); + if (err) + break; + dprint(PRINT_DEBUG, "IOCTL SET: got arg %d\n", val); + if (val < 0 || val > PRINT_MAX) { + err = -EINVAL; + break; + } + set_debug_mask(val); + break; + + /* add non-debugging fist ioctl's here */ + + case UNIONFS_IOCTL_BRANCH_COUNT: + if (!capable(CAP_SYS_ADMIN)) { + err = -EACCES; + goto out; + } + err = unionfs_ioctl_branchcount(file, cmd, arg); + break; + + case UNIONFS_IOCTL_INCGEN: + if (!capable(CAP_SYS_ADMIN)) { + err = -EACCES; + goto out; + } + err = unionfs_ioctl_incgen(file, cmd, arg); + break; + + case UNIONFS_IOCTL_ADDBRANCH: + if (!capable(CAP_SYS_ADMIN)) { + err = -EACCES; + goto out; + } + err = + unionfs_ioctl_addbranch(file->f_dentry->d_inode, cmd, arg); + break; + + case UNIONFS_IOCTL_RDWRBRANCH: + if (!capable(CAP_SYS_ADMIN)) { + err = -EACCES; + goto out; + } + err = + unionfs_ioctl_rdwrbranch(file->f_dentry->d_inode, cmd, arg); + break; + + case UNIONFS_IOCTL_QUERYFILE: + /* XXX: This should take the file. */ + err = unionfs_ioctl_queryfile(file, cmd, arg); + break; + + default: + hidden_file = ftohf(file); + + err = -ENOTTY; + if (!hidden_file || !hidden_file->f_op) + goto out; + if (hidden_file->f_op->unlocked_ioctl) { + err = + hidden_file->f_op->unlocked_ioctl(hidden_file, cmd, + arg); + } else if (hidden_file->f_op->ioctl) { + lock_kernel(); + err = + hidden_file->f_op->ioctl(hidden_file->f_dentry-> + d_inode, hidden_file, cmd, + arg); + unlock_kernel(); + } + } /* end of outer switch statement */ + + out: + print_exit_status((int)err); + return err; +} + +int unionfs_flush(struct file *file, fl_owner_t id) +{ + int err = 0; /* assume ok (see open.c:close_fp) */ + struct file *hidden_file = NULL; + int bindex, bstart, bend; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 1))) + goto out; + if (!atomic_dec_and_test + (&itopd(file->f_dentry->d_inode)->uii_totalopens)) + goto out; + + lock_dentry(file->f_dentry); + + bstart = fbstart(file); + bend = fbend(file); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_file = ftohf_index(file, bindex); + + if (hidden_file && hidden_file->f_op + && hidden_file->f_op->flush) { + err = hidden_file->f_op->flush(hidden_file, id); + if (err) + goto out_lock; + /* This was earlier done in the unlink_all function in unlink.c */ + /* if there are no more references to the dentry, dput it */ + if (d_deleted(file->f_dentry)) { + DPUT(dtohd_index(file->f_dentry, bindex)); + set_dtohd_index(file->f_dentry, bindex, NULL); + } + } + + } + + out_lock: + unlock_dentry(file->f_dentry); + out: + print_exit_status(err); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/copyup.c @@ -0,0 +1,738 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York* + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: copyup.c,v 1.78 2006/10/31 18:05:33 yiannos Exp $ + */ + +#include "unionfs.h" + +/*Not Working Yet*/ +static int copyup_xattrs(struct dentry *old_hidden_dentry, + struct dentry *new_hidden_dentry) +{ + int err = 0; + ssize_t list_size = -1; + char *name_list = NULL; + char *attr_value = NULL; + char *name_list_orig = NULL; + + print_entry_location(); + + list_size = vfs_listxattr(old_hidden_dentry, NULL, NULL, 0, NULL); + + if (list_size <= 0) { + err = list_size; + goto out; + } + + name_list = xattr_alloc(list_size + 1, XATTR_LIST_MAX); + if (!name_list || IS_ERR(name_list)) { + err = PTR_ERR(name_list); + goto out; + } + list_size = vfs_listxattr(old_hidden_dentry, NULL, name_list, + list_size, NULL); + attr_value = xattr_alloc(XATTR_SIZE_MAX, XATTR_SIZE_MAX); + if (!attr_value || IS_ERR(attr_value)) { + err = PTR_ERR(name_list); + goto out; + } + name_list_orig = name_list; + while (*name_list) { + ssize_t size; + + //We need to lock here since vfs_getxattr doesn't lock for us. + mutex_lock(&old_hidden_dentry->d_inode->i_mutex); + size = vfs_getxattr(old_hidden_dentry, NULL, name_list, + attr_value, XATTR_SIZE_MAX, NULL); + mutex_unlock(&old_hidden_dentry->d_inode->i_mutex); + if (size < 0) { + err = size; + goto out; + } + + if (size > XATTR_SIZE_MAX) { + err = -E2BIG; + goto out; + } + //We don't need to lock here since vfs_setxattr does it for us. + err = vfs_setxattr(new_hidden_dentry, NULL, name_list, + attr_value, size, 0, NULL); + + if (err < 0) + goto out; + name_list += strlen(name_list) + 1; + } + out: + name_list = name_list_orig; + + if (name_list) + xattr_free(name_list, list_size + 1); + if (attr_value) + xattr_free(attr_value, XATTR_SIZE_MAX); + /* It is no big deal if this fails, we just roll with the punches. */ + if (err == -ENOTSUPP || err == -EOPNOTSUPP) + err = 0; + return err; +} + +/* Determine the mode based on the copyup flags, and the existing dentry. */ +static int copyup_permissions(struct super_block *sb, + struct dentry *old_hidden_dentry, + struct dentry *new_hidden_dentry) +{ + struct iattr newattrs; + int err; + + print_entry_location(); + + newattrs.ia_atime = old_hidden_dentry->d_inode->i_atime; + newattrs.ia_mtime = old_hidden_dentry->d_inode->i_mtime; + newattrs.ia_ctime = old_hidden_dentry->d_inode->i_ctime; + newattrs.ia_valid = ATTR_CTIME | ATTR_ATIME | ATTR_MTIME | + ATTR_ATIME_SET | ATTR_MTIME_SET; + /* original mode of old file */ + newattrs.ia_mode = old_hidden_dentry->d_inode->i_mode; + newattrs.ia_gid = old_hidden_dentry->d_inode->i_gid; + newattrs.ia_uid = old_hidden_dentry->d_inode->i_uid; + newattrs.ia_valid |= ATTR_FORCE | ATTR_GID | ATTR_UID | ATTR_MODE; + if (newattrs.ia_valid & ATTR_MODE) { + newattrs.ia_mode = + (newattrs.ia_mode & S_IALLUGO) | (old_hidden_dentry-> + d_inode-> + i_mode & ~S_IALLUGO); + } + + err = notify_change(new_hidden_dentry, NULL, &newattrs); + + print_exit_status(err); + return err; +} + +int copyup_dentry(struct inode *dir, struct dentry *dentry, + int bstart, int new_bindex, + struct file **copyup_file, loff_t len) +{ + return copyup_named_dentry(dir, dentry, bstart, new_bindex, + dentry->d_name.name, + dentry->d_name.len, copyup_file, len); +} + +int copyup_named_dentry(struct inode *dir, struct dentry *dentry, + int bstart, int new_bindex, const char *name, + int namelen, struct file **copyup_file, loff_t len) +{ + struct dentry *new_hidden_dentry; + struct dentry *old_hidden_dentry = NULL; + struct super_block *sb; + struct file *input_file = NULL; + struct file *output_file = NULL; + struct sioq_args args; + ssize_t read_bytes, write_bytes; + mm_segment_t old_fs; + int err = 0; + char *buf; + int old_bindex; + int got_branch_input = -1; + int got_branch_output = -1; + int old_bstart; + int old_bend; + int old_mode; + loff_t size = len; + struct dentry *new_hidden_parent_dentry = NULL; + mm_segment_t oldfs; + char *symbuf = NULL; + + print_entry_location(); + verify_locked(dentry); + print_dentry("IN: copyup_named_dentry", dentry); + + old_bindex = bstart; + old_bstart = dbstart(dentry); + old_bend = dbend(dentry); + + BUG_ON(new_bindex < 0); + BUG_ON(new_bindex >= old_bindex); + + sb = dir->i_sb; + + unionfs_read_lock(sb); + + if ((err = is_robranch_super(sb, new_bindex))) + goto out; + + /* Create the directory structure above this dentry. */ + new_hidden_dentry = create_parents_named(dir, dentry, name, new_bindex); + if (IS_ERR(new_hidden_dentry)) { + err = PTR_ERR(new_hidden_dentry); + goto out; + } + + print_dentry("Copyup Object", new_hidden_dentry); + + /* Now we actually create the object. */ + old_hidden_dentry = dtohd_index(dentry, old_bindex); + DGET(old_hidden_dentry); + + old_mode = old_hidden_dentry->d_inode->i_mode; + + /* For symlinks, we must read the link before we lock the directory. */ + if (S_ISLNK(old_mode)) { + + symbuf = KMALLOC(PATH_MAX, GFP_KERNEL); + if (!symbuf) { + err = -ENOMEM; + goto copyup_readlink_err; + } + + oldfs = get_fs(); + set_fs(KERNEL_DS); + err = + old_hidden_dentry->d_inode->i_op-> + readlink(old_hidden_dentry, (char __user *)symbuf, + PATH_MAX); + set_fs(oldfs); + if (err < 0) + goto copyup_readlink_err; + symbuf[err] = '\0'; + } + + /* Now we lock the parent, and create the object in the new branch. */ + new_hidden_parent_dentry = lock_parent(new_hidden_dentry); + if (S_ISDIR(old_mode)) { + args.mkdir.parent = new_hidden_parent_dentry->d_inode; + args.mkdir.dentry = new_hidden_dentry; + args.mkdir.mode = old_mode; /*S_IRWXU*/ + run_sioq(__unionfs_mkdir, &args); + err = args.err; + } else if (S_ISLNK(old_mode)) { + args.symlink.parent = new_hidden_parent_dentry->d_inode; + args.symlink.dentry = new_hidden_dentry; + args.symlink.symbuf = symbuf; + args.symlink.mode = old_mode; + run_sioq(__unionfs_symlink, &args); + err = args.err; + } else if (S_ISBLK(old_mode) + || S_ISCHR(old_mode) + || S_ISFIFO(old_mode) + || S_ISSOCK(old_mode)) { + args.mknod.parent = new_hidden_parent_dentry->d_inode; + args.mknod.dentry = new_hidden_dentry; + args.mknod.mode = old_mode; + args.mknod.dev = old_hidden_dentry->d_inode->i_rdev; + run_sioq(__unionfs_mknod, &args); + err = args.err; + } else if (S_ISREG(old_mode)) { + args.create.parent = new_hidden_parent_dentry->d_inode; + args.create.dentry = new_hidden_dentry; + args.create.mode = old_mode; + args.create.nd = NULL; + run_sioq(__unionfs_create, &args); + err = args.err; + } else { + printk(KERN_ERR "Unknown inode type %d\n", + old_hidden_dentry->d_inode->i_mode); + BUG(); + } + + copyup_readlink_err: + KFREE(symbuf); + if (err) { + /* get rid of the hidden dentry and all its traces */ + DPUT(new_hidden_dentry); + set_dtohd_index(dentry, new_bindex, NULL); + set_dbstart(dentry, old_bstart); + set_dbend(dentry, old_bend); + goto out_dir; + } +#ifdef UNIONFS_IMAP + if (stopd(sb)->usi_persistent) { + err = write_uin(dentry->d_sb, dentry->d_inode->i_ino, + new_bindex, new_hidden_dentry->d_inode->i_ino); + if (err) + goto out_dir; + } +#endif + /* We actually copyup the file here. */ + if (S_ISREG(old_hidden_dentry->d_inode->i_mode)) { + mntget(stohiddenmnt_index(sb, old_bindex)); + branchget(sb, old_bindex); + got_branch_input = old_bindex; + input_file = + DENTRY_OPEN(old_hidden_dentry, + stohiddenmnt_index(sb, old_bindex), + O_RDONLY | O_LARGEFILE); + if (IS_ERR(input_file)) { + err = PTR_ERR(input_file); + goto out_dir; + } + if (!input_file->f_op || !input_file->f_op->read) { + err = -EINVAL; + goto out_dir; + } + + /* copy the new file */ + DGET(new_hidden_dentry); + mntget(stohiddenmnt_index(sb, new_bindex)); + branchget(sb, new_bindex); + got_branch_output = new_bindex; + output_file = + DENTRY_OPEN(new_hidden_dentry, + stohiddenmnt_index(sb, new_bindex), + O_WRONLY | O_LARGEFILE); + if (IS_ERR(output_file)) { + err = PTR_ERR(output_file); + goto out_dir; + } + if (!output_file->f_op || !output_file->f_op->write) { + err = -EINVAL; + goto out_dir; + } + + /* allocating a buffer */ + buf = (char *)KMALLOC(PAGE_SIZE, GFP_KERNEL); + if (!buf) { + err = -ENOMEM; + goto out_dir; + } + + /* now read PAGE_SIZE bytes from offset 0 in a loop */ + old_fs = get_fs(); + + input_file->f_pos = 0; + output_file->f_pos = 0; + + err = 0; // reset error just in case + set_fs(KERNEL_DS); + do { + if (len >= PAGE_SIZE) + size = PAGE_SIZE; + else if ((len < PAGE_SIZE) && (len > 0)) + size = len; + + len -= PAGE_SIZE; + + read_bytes = + input_file->f_op->read(input_file, + (char __user *)buf, size, + &input_file->f_pos); + if (read_bytes <= 0) { + err = read_bytes; + break; + } + + write_bytes = + output_file->f_op->write(output_file, + (char __user *)buf, + read_bytes, + &output_file->f_pos); + if (write_bytes < 0 || (write_bytes < read_bytes)) { + err = write_bytes; + break; + } + } while ((read_bytes > 0) && (len > 0)); + set_fs(old_fs); + KFREE(buf); +#ifdef UNIONFS_MMAP + /* SP: Now that we copied up the file, have to sync its data + * as otherwise when we do a read_cache_page(), we'll possibly + * read crap. + * + * another posisble solution would be in the address op code + * would be to check the "lower" page to see if its dirty, + * and if it's dirty, use it directl + */ + if (!err) { + err = + output_file->f_op->fsync(output_file, + new_hidden_dentry, 0); + } +#endif + if (err) { + /* copyup failed, because we ran out of space or quota, + * or something else happened so let's unlink; we don't + * really care about the return value of vfs_unlink */ + vfs_unlink(new_hidden_parent_dentry->d_inode, + new_hidden_dentry, NULL); + + goto out_dir; + } + } + + /* Set permissions. */ + if ((err = + copyup_permissions(sb, old_hidden_dentry, new_hidden_dentry))) + goto out_dir; + /* Selinux uses extended attributes for permissions. */ + if ((err = copyup_xattrs(old_hidden_dentry, new_hidden_dentry))) + goto out_dir; + + /* do not allow files getting deleted to be reinterposed */ + if (!d_deleted(dentry)) + unionfs_reinterpose(dentry); + + out_dir: + if (new_hidden_parent_dentry) + unlock_dir(new_hidden_parent_dentry); + + out: + if (input_file && !IS_ERR(input_file)) { + fput(input_file); + } else { + /* since input file was not opened, we need to explicitly + * dput the old_hidden_dentry + */ + DPUT(old_hidden_dentry); + } + + /* in any case, we have to branchput */ + if (got_branch_input >= 0) + branchput(sb, got_branch_input); + + if (output_file) { + if (copyup_file && !err) { + *copyup_file = output_file; + } else { + /* close the file if there was no error, or if we ran + * out of space in which case we unlinked the file */ + if (!IS_ERR(output_file)) + fput(output_file); + branchput(sb, got_branch_output); + } + } + + unionfs_read_unlock(sb); + + print_dentry("OUT: copyup_dentry", dentry); + print_inode("OUT: copyup_dentry", dentry->d_inode); + + print_exit_status(err); + return err; +} + +/* This function creates a copy of a file represented by 'file' which currently + * resides in branch 'bstart' to branch 'new_bindex. The copy will be named + * "name". */ +int copyup_named_file(struct inode *dir, struct file *file, char *name, + int bstart, int new_bindex, loff_t len) +{ + int err = 0; + struct file *output_file = NULL; + + print_entry_location(); + + err = copyup_named_dentry(dir, file->f_dentry, bstart, + new_bindex, name, strlen(name), &output_file, + len); + if (!err) { + fbstart(file) = new_bindex; + set_ftohf_index(file, new_bindex, output_file); + } + + print_exit_status(err); + return err; +} + +/* This function creates a copy of a file represented by 'file' which currently + * resides in branch 'bstart' to branch 'new_bindex. + */ +int copyup_file(struct inode *dir, struct file *file, int bstart, + int new_bindex, loff_t len) +{ + int err = 0; + struct file *output_file = NULL; + + print_entry_location(); + + err = copyup_dentry(dir, file->f_dentry, bstart, new_bindex, + &output_file, len); + if (!err) { + fbstart(file) = new_bindex; + set_ftohf_index(file, new_bindex, output_file); + } + + print_exit_status(err); + return err; +} + +/* This function replicates the directory structure upto given dentry + * in the bindex branch. Can create directory structure recursively to the right + * also. + */ +struct dentry *create_parents(struct inode *dir, struct dentry *dentry, + int bindex) +{ + struct dentry *hidden_dentry; + + print_entry_location(); + hidden_dentry = + create_parents_named(dir, dentry, dentry->d_name.name, bindex); + print_exit_location(); + + return (hidden_dentry); +} + +/* This function replicates the directory structure upto given dentry + * in the bindex branch. */ +struct dentry *create_parents_named(struct inode *dir, struct dentry *dentry, + const char *name, int bindex) +{ + int err; + struct dentry *child_dentry; + struct dentry *parent_dentry; + struct dentry *hidden_parent_dentry = NULL; + struct dentry *hidden_dentry = NULL; + struct sioq_args args; + const char *childname; + unsigned int childnamelen; + + int old_kmalloc_size; + int kmalloc_size; + int num_dentry; + int count; + + int old_bstart; + int old_bend; + struct dentry **path = NULL; + struct dentry **tmp_path; + struct super_block *sb; +#ifdef UNIONFS_IMAP + int persistent; +#endif + print_entry_location(); + + verify_locked(dentry); + + /* There is no sense allocating any less than the minimum. */ + kmalloc_size = 128; + num_dentry = kmalloc_size / sizeof(struct dentry *); + + if ((err = is_robranch_super(dir->i_sb, bindex))) { + hidden_dentry = ERR_PTR(err); + goto out; + } + + print_dentry("IN: create_parents_named", dentry); + dprint(PRINT_DEBUG, "name = %s\n", name); + + old_bstart = dbstart(dentry); + old_bend = dbend(dentry); + + hidden_dentry = ERR_PTR(-ENOMEM); + path = (struct dentry **)KZALLOC(kmalloc_size, GFP_KERNEL); + if (!path) + goto out; + + /* assume the negative dentry of unionfs as the parent dentry */ + parent_dentry = dentry; + + count = 0; + /* This loop finds the first parent that exists in the given branch. + * We start building the directory structure from there. At the end + * of the loop, the following should hold: + * child_dentry is the first nonexistent child + * parent_dentry is the first existent parent + * path[0] is the = deepest child + * path[count] is the first child to create + */ + do { + child_dentry = parent_dentry; + + /* find the parent directory dentry in unionfs */ + parent_dentry = child_dentry->d_parent; + lock_dentry(parent_dentry); + + /* find out the hidden_parent_dentry in the given branch */ + hidden_parent_dentry = dtohd_index(parent_dentry, bindex); + + /* store the child dentry */ + path[count++] = child_dentry; + if (count == num_dentry) { + old_kmalloc_size = kmalloc_size; + kmalloc_size *= 2; + num_dentry = kmalloc_size / sizeof(struct dentry *); + + tmp_path = + (struct dentry **)KZALLOC(kmalloc_size, GFP_KERNEL); + if (!tmp_path) { + hidden_dentry = ERR_PTR(-ENOMEM); + goto out; + } + memcpy(tmp_path, path, old_kmalloc_size); + KFREE(path); + path = tmp_path; + tmp_path = NULL; + } + + } while (!hidden_parent_dentry); + count--; + + sb = dentry->d_sb; +#ifdef UNIONFS_IMAP + persistent = stopd(sb)->usi_persistent; +#endif + /* This is basically while(child_dentry != dentry). This loop is + * horrible to follow and should be replaced with cleaner code. */ + while (1) { + // get hidden parent dir in the current branch + hidden_parent_dentry = dtohd_index(parent_dentry, bindex); + unlock_dentry(parent_dentry); + + // init the values to lookup + childname = child_dentry->d_name.name; + childnamelen = child_dentry->d_name.len; + + if (child_dentry != dentry) { + // lookup child in the underlying file system + hidden_dentry = + LOOKUP_ONE_LEN(childname, hidden_parent_dentry, + childnamelen); + if (IS_ERR(hidden_dentry)) + goto out; + } else { + int loop_start; + int loop_end; + int new_bstart = -1; + int new_bend = -1; + int i; + + /* is the name a whiteout of the childname ? */ + //lookup the whiteout child in the underlying file system + hidden_dentry = + LOOKUP_ONE_LEN(name, hidden_parent_dentry, + strlen(name)); + if (IS_ERR(hidden_dentry)) + goto out; + + /* Replace the current dentry (if any) with the new one. */ + DPUT(dtohd_index(dentry, bindex)); + set_dtohd_index(dentry, bindex, hidden_dentry); + + loop_start = + (old_bstart < bindex) ? old_bstart : bindex; + loop_end = (old_bend > bindex) ? old_bend : bindex; + + /* This loop sets the bstart and bend for the new + * dentry by traversing from left to right. + * It also dputs all negative dentries except + * bindex (the newly looked dentry + */ + for (i = loop_start; i <= loop_end; i++) { + if (!dtohd_index(dentry, i)) + continue; + + if (i == bindex) { + new_bend = i; + if (new_bstart < 0) + new_bstart = i; + continue; + } + + if (!dtohd_index(dentry, i)->d_inode) { + DPUT(dtohd_index(dentry, i)); + set_dtohd_index(dentry, i, NULL); + } else { + if (new_bstart < 0) + new_bstart = i; + new_bend = i; + } + } + + if (new_bstart < 0) + new_bstart = bindex; + if (new_bend < 0) + new_bend = bindex; + set_dbstart(dentry, new_bstart); + set_dbend(dentry, new_bend); + break; + } + + if (hidden_dentry->d_inode) { + /* since this already exists we dput to avoid + * multiple references on the same dentry */ + DPUT(hidden_dentry); + } else { + + /* its a negative dentry, create a new dir */ + hidden_parent_dentry = lock_parent(hidden_dentry); + args.mkdir.parent = hidden_parent_dentry->d_inode; + args.mkdir.dentry = hidden_dentry; + args.mkdir.mode = child_dentry->d_inode->i_mode; + run_sioq(__unionfs_mkdir, &args); + err = args.err; + if (!err) + err = copyup_permissions + (dir->i_sb, child_dentry, hidden_dentry); + unlock_dir(hidden_parent_dentry); + if (err) { + DPUT(hidden_dentry); + hidden_dentry = ERR_PTR(err); + goto out; + } +#ifdef UNIONFS_IMAP + if (persistent) { + err = write_uin + (sb, child_dentry->d_inode->i_ino, + bindex, hidden_dentry->d_inode->i_ino); + if (err) { + DPUT(hidden_dentry); + hidden_dentry = ERR_PTR(err); + goto out; + } + } +#endif + set_itohi_index(child_dentry->d_inode, bindex, + IGRAB(hidden_dentry->d_inode)); + if (ibstart(child_dentry->d_inode) > bindex) + ibstart(child_dentry->d_inode) = bindex; + if (ibend(child_dentry->d_inode) < bindex) + ibend(child_dentry->d_inode) = bindex; + + set_dtohd_index(child_dentry, bindex, hidden_dentry); + if (dbstart(child_dentry) > bindex) + set_dbstart(child_dentry, bindex); + if (dbend(child_dentry) < bindex) + set_dbend(child_dentry, bindex); + } + + parent_dentry = child_dentry; + child_dentry = path[--count]; + } + out: + KFREE(path); + print_dentry("OUT: create_parents_named", dentry); + print_exit_pointer(hidden_dentry); + return hidden_dentry; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/dentry.c @@ -0,0 +1,282 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: dentry.c,v 1.77 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +/* declarations added for "sparse" */ +extern int unionfs_d_revalidate_wrap(struct dentry *dentry, + struct nameidata *nd); +extern void unionfs_d_release(struct dentry *dentry); +extern void unionfs_d_iput(struct dentry *dentry, struct inode *inode); + +/* + * THIS IS A BOOLEAN FUNCTION: returns 1 if valid, 0 otherwise. + */ +int unionfs_d_revalidate(struct dentry *dentry, struct nameidata *nd) +{ + int valid = 1; /* default is valid (1); invalid is 0. */ + struct dentry *hidden_dentry; + int bindex, bstart, bend; + int sbgen, dgen; + int positive = 0; + int locked = 0; + int restart = 0; + int interpose_flag; + + struct nameidata lowernd; + + if(nd) + memcpy(&lowernd, nd, sizeof(struct nameidata)); + else + memset(&lowernd, 0, sizeof(struct nameidata)); + + print_util_entry_location(); + + restart: + verify_locked(dentry); + + /* if the dentry is unhashed, do NOT revalidate */ + if (d_deleted(dentry)) { + dprint(PRINT_DEBUG, "unhashed dentry being revalidated: %*s\n", + dentry->d_name.len, dentry->d_name.name); + goto out; + } + + BUG_ON(dbstart(dentry) == -1); + if (dentry->d_inode) + positive = 1; + dgen = atomic_read(&dtopd(dentry)->udi_generation); + sbgen = atomic_read(&stopd(dentry->d_sb)->usi_generation); + /* If we are working on an unconnected dentry, then there is no + * revalidation to be done, because this file does not exist within the + * namespace, and Unionfs operates on the namespace, not data. + */ + if (sbgen != dgen) { + struct dentry *result; + int pdgen; + + unionfs_read_lock(dentry->d_sb); + locked = 1; + + /* The root entry should always be valid */ + BUG_ON(IS_ROOT(dentry)); + + /* We can't work correctly if our parent isn't valid. */ + pdgen = atomic_read(&dtopd(dentry->d_parent)->udi_generation); + if (!restart && (pdgen != sbgen)) { + unionfs_read_unlock(dentry->d_sb); + locked = 0; + /* We must be locked before our parent. */ + if (! + (dentry->d_parent->d_op-> + d_revalidate(dentry->d_parent, nd))) { + valid = 0; + goto out; + } + restart = 1; + goto restart; + } + BUG_ON(pdgen != sbgen); + + /* Free the pointers for our inodes and this dentry. */ + bstart = dbstart(dentry); + bend = dbend(dentry); + if (bstart >= 0) { + struct dentry *hidden_dentry; + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = + dtohd_index_nocheck(dentry, bindex); + if (!hidden_dentry) + continue; + DPUT(hidden_dentry); + } + } + set_dbstart(dentry, -1); + set_dbend(dentry, -1); + + interpose_flag = INTERPOSE_REVAL_NEG; + if (positive) { + interpose_flag = INTERPOSE_REVAL; + mutex_lock(&dentry->d_inode->i_mutex); + bstart = ibstart(dentry->d_inode); + bend = ibend(dentry->d_inode); + if (bstart >= 0) { + struct inode *hidden_inode; + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_inode = + itohi_index(dentry->d_inode, + bindex); + if (!hidden_inode) + continue; + IPUT(hidden_inode); + } + } + KFREE(itohi_ptr(dentry->d_inode)); + itohi_ptr(dentry->d_inode) = NULL; + ibstart(dentry->d_inode) = -1; + ibend(dentry->d_inode) = -1; + mutex_unlock(&dentry->d_inode->i_mutex); + } + + result = unionfs_lookup_backend(dentry, &lowernd, interpose_flag); + if (result) { + if (IS_ERR(result)) { + valid = 0; + goto out; + } + /* current unionfs_lookup_backend() doesn't return + a valid dentry */ + DPUT(dentry); + dentry = result; + } + + if (positive && itopd(dentry->d_inode)->uii_stale) { + make_stale_inode(dentry->d_inode); + d_drop(dentry); + valid = 0; + goto out; + } + goto out; + } + + /* The revalidation must occur across all branches */ + bstart = dbstart(dentry); + bend = dbend(dentry); + BUG_ON(bstart == -1); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry || !hidden_dentry->d_op + || !hidden_dentry->d_op->d_revalidate) + continue; + + if (!hidden_dentry->d_op->d_revalidate(hidden_dentry, nd)) + valid = 0; + } + + if (!dentry->d_inode) + valid = 0; + if (valid) + fist_copy_attr_all(dentry->d_inode, itohi(dentry->d_inode)); + + out: + if (locked) + unionfs_read_unlock(dentry->d_sb); + print_dentry("revalidate out", dentry); + print_util_exit_status(valid); + return valid; +} + +int unionfs_d_revalidate_wrap(struct dentry *dentry, struct nameidata *nd) +{ + int err; + + print_entry_location(); + lock_dentry(dentry); + + err = unionfs_d_revalidate(dentry, nd); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +void unionfs_d_release(struct dentry *dentry) +{ + struct dentry *hidden_dentry; + int bindex, bstart, bend; + + print_entry_location(); + /* There is no reason to lock the dentry, because we have the only + * reference, but the printing functions verify that we have a lock + * on the dentry before calling dbstart, etc. */ + lock_dentry(dentry); + print_dentry_nocheck("unionfs_d_release IN dentry", dentry); + + /* this could be a negative dentry, so check first */ + if (!dtopd(dentry)) { + dprint(PRINT_DEBUG, "dentry without private data: %*s", + dentry->d_name.len, dentry->d_name.name); + goto out; + } else if (dbstart(dentry) < 0) { + /* this is due to a failed lookup */ + /* the failed lookup has a dtohd_ptr set to null, + but this is a better check */ + dprint(PRINT_DEBUG, "dentry without hidden dentries : %*s", + dentry->d_name.len, dentry->d_name.name); + goto out_free; + } + + /* Release all the hidden dentries */ + bstart = dbstart(dentry); + bend = dbend(dentry); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + DPUT(hidden_dentry); + set_dtohd_index(dentry, bindex, NULL); + } + /* free private data (unionfs_dentry_info) here */ + KFREE(dtohd_ptr(dentry)); + dtohd_ptr(dentry) = NULL; + out_free: + /* No need to unlock it, because it is disappeared. */ +#ifdef TRACKLOCK + printk("DESTROYLOCK:%p\n", dentry); +#endif + free_dentry_private_data(dtopd(dentry)); + dtopd_lhs(dentry) = NULL; /* just to be safe */ + out: + print_exit_location(); +} + +/* + * we don't really need unionfs_d_iput, because dentry_iput will call iput() if + * unionfs_d_iput is not defined. We left this implemented for ease of + * tracing/debugging. + */ +void unionfs_d_iput(struct dentry *dentry, struct inode *inode) +{ + print_entry_location(); + IPUT(inode); + print_exit_location(); +} + +struct dentry_operations unionfs_dops = { + .d_revalidate = unionfs_d_revalidate_wrap, + .d_release = unionfs_d_release, + .d_iput = unionfs_d_iput, +}; + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/dirfops.c @@ -0,0 +1,330 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: dirfops.c,v 1.25 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +/* Make sure our rdstate is playing by the rules. */ +static void verify_rdstate_offset(struct unionfs_dir_state *rdstate) +{ + BUG_ON(rdstate->uds_offset >= DIREOF); + BUG_ON(rdstate->uds_cookie >= MAXRDCOOKIE); +} + +struct unionfs_getdents_callback { + struct unionfs_dir_state *rdstate; + void *dirent; + int entries_written; + int filldir_called; + int filldir_error; + filldir_t filldir; + struct super_block *sb; +}; + +/* copied from generic filldir in fs/readir.c */ +static int unionfs_filldir(void *dirent, const char *name, int namelen, + loff_t offset, u64 ino, unsigned int d_type) +{ + struct unionfs_getdents_callback *buf = + (struct unionfs_getdents_callback *)dirent; + struct filldir_node *found = NULL; + int err = 0; + int is_wh_entry = 0; + + dprint(PRINT_DEBUG, "unionfs_filldir name=%*s\n", namelen, name); + + buf->filldir_called++; + + if ((namelen > WHLEN) && !strncmp(name, WHPFX, WHLEN)) { + name += WHLEN; + namelen -= WHLEN; + is_wh_entry = 1; + } + + found = find_filldir_node(buf->rdstate, name, namelen); + + if (found) + goto out; + + /* if 'name' isn't a whiteout filldir it. */ + if (!is_wh_entry) { + off_t pos = rdstate2offset(buf->rdstate); + ino_t unionfs_ino = ino; +#ifdef UNIONFS_IMAP + if (stopd(buf->sb)->usi_persistent) + err = read_uin(buf->sb, buf->rdstate->uds_bindex, + ino, O_CREAT, &unionfs_ino); +#endif + if (!err) { + err = buf->filldir(buf->dirent, name, namelen, pos, + unionfs_ino, d_type); + buf->rdstate->uds_offset++; + verify_rdstate_offset(buf->rdstate); + } + } + /* If we did fill it, stuff it in our hash, otherwise return an error */ + if (err) { + buf->filldir_error = err; + goto out; + } + buf->entries_written++; + if ((err = add_filldir_node(buf->rdstate, name, namelen, + buf->rdstate->uds_bindex, is_wh_entry))) + buf->filldir_error = err; + + out: + return err; +} + +static int unionfs_readdir(struct file *file, void *dirent, filldir_t filldir) +{ + int err = 0; + struct file *hidden_file = NULL; + struct inode *inode = NULL; + struct unionfs_getdents_callback buf; + struct unionfs_dir_state *uds; + int bend; + loff_t offset; + + print_entry("file = %p, pos = %llx", file, file->f_pos); + + print_file("In unionfs_readdir()", file); + + if ((err = unionfs_file_revalidate(file, 0))) + goto out; + + inode = file->f_dentry->d_inode; + checkinode(inode, "unionfs_readdir"); + + uds = ftopd(file)->rdstate; + if (!uds) { + if (file->f_pos == DIREOF) { + goto out; + } else if (file->f_pos > 0) { + uds = find_rdstate(inode, file->f_pos); + if (!uds) { + err = -ESTALE; + goto out; + } + ftopd(file)->rdstate = uds; + } else { + init_rdstate(file); + uds = ftopd(file)->rdstate; + } + } + bend = fbend(file); + + while (uds->uds_bindex <= bend) { + hidden_file = ftohf_index(file, uds->uds_bindex); + if (!hidden_file) { + dprint(PRINT_DEBUG, + "Incremented bindex to %d of %d," + " because hidden file is NULL.\n", + uds->uds_bindex, bend); + uds->uds_bindex++; + uds->uds_dirpos = 0; + continue; + } + + /* prepare callback buffer */ + buf.filldir_called = 0; + buf.filldir_error = 0; + buf.entries_written = 0; + buf.dirent = dirent; + buf.filldir = filldir; + buf.rdstate = uds; + buf.sb = inode->i_sb; + + /* Read starting from where we last left off. */ + offset = vfs_llseek(hidden_file, uds->uds_dirpos, 0); + if (offset < 0) { + err = offset; + goto out; + } + dprint(PRINT_DEBUG, "calling readdir for %d.%lld (offset = %lld)\n", + uds->uds_bindex, uds->uds_dirpos, offset); + err = vfs_readdir(hidden_file, unionfs_filldir, (void *)&buf); + dprint(PRINT_DEBUG, + "readdir on %d.%lld = %d (entries written %d, filldir called %d)\n", + uds->uds_bindex, (long long)uds->uds_dirpos, err, + buf.entries_written, buf.filldir_called); + /* Save the position for when we continue. */ + + offset = vfs_llseek(hidden_file, 0, 1); + if (offset < 0) { + err = offset; + goto out; + } + uds->uds_dirpos = offset; + + /* Copy the atime. */ + fist_copy_attr_atime(inode, hidden_file->f_dentry->d_inode); + + if (err < 0) { + goto out; + } + + if (buf.filldir_error) { + break; + } + + if (!buf.entries_written) { + uds->uds_bindex++; + uds->uds_dirpos = 0; + } + } + + if (!buf.filldir_error && uds->uds_bindex >= bend) { + dprint(PRINT_DEBUG, + "Discarding rdstate because readdir is over (hashsize = %d)\n", + uds->uds_hashentries); + /* Save the number of hash entries for next time. */ + itopd(inode)->uii_hashsize = uds->uds_hashentries; + free_rdstate(uds); + ftopd(file)->rdstate = NULL; + file->f_pos = DIREOF; + } else { + file->f_pos = rdstate2offset(uds); + dprint(PRINT_DEBUG, "rdstate now has a cookie of %u (err = %d)\n", + uds->uds_cookie, err); + } + + out: + checkinode(inode, "post unionfs_readdir"); + print_exit_status(err); + return err; +} + +/* This is not meant to be a generic repositioning function. If you do + * things that aren't supported, then we return EINVAL. + * + * What is allowed: + * (1) seeking to the same position that you are currently at + * This really has no effect, but returns where you are. + * (2) seeking to the end of the file, if you've read everything + * This really has no effect, but returns where you are. + * (3) seeking to the beginning of the file + * This throws out all state, and lets you begin again. + */ +static loff_t unionfs_dir_llseek(struct file *file, loff_t offset, int origin) +{ + struct unionfs_dir_state *rdstate; + loff_t err; + + print_entry(" file=%p, offset=0x%llx, origin = %d", file, offset, + origin); + + if ((err = unionfs_file_revalidate(file, 0))) + goto out; + + rdstate = ftopd(file)->rdstate; + + /* We let users seek to their current position, but not anywhere else. */ + if (!offset) { + switch (origin) { + case SEEK_SET: + if (rdstate) { + free_rdstate(rdstate); + ftopd(file)->rdstate = NULL; + } + init_rdstate(file); + err = 0; + break; + case SEEK_CUR: + if (file->f_pos) { + if (file->f_pos == DIREOF) + err = DIREOF; + else + BUG_ON(file->f_pos != + rdstate2offset(rdstate)); + err = file->f_pos; + } else { + err = 0; + } + break; + case SEEK_END: + /* Unsupported, because we would break everything. */ + err = -EINVAL; + break; + } + } else { + switch (origin) { + case SEEK_SET: + if (rdstate) { + if (offset == rdstate2offset(rdstate)) { + err = offset; + } else if (file->f_pos == DIREOF) { + err = DIREOF; + } else { + err = -EINVAL; + } + } else { + if ((rdstate = + find_rdstate(file->f_dentry->d_inode, + offset))) { + ftopd(file)->rdstate = rdstate; + err = rdstate->uds_offset; + } else { + err = -EINVAL; + } + } + break; + case SEEK_CUR: + case SEEK_END: + /* Unsupported, because we would break everything. */ + err = -EINVAL; + break; + } + } + + out: + print_exit_status((int)err); + return err; +} + +/* Trimmed directory options, we shouldn't pass everything down since + * we don't want to operate on partial directories. + */ +struct file_operations unionfs_dir_fops = { + .llseek = unionfs_dir_llseek, + .read = generic_read_dir, + .readdir = unionfs_readdir, + .unlocked_ioctl = unionfs_ioctl, + .open = unionfs_open, + .release = unionfs_file_release, + .flush = unionfs_flush, +}; + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/dirhelper.c @@ -0,0 +1,309 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: dirhelper.c,v 1.32 2006/10/31 18:05:33 yiannos Exp $ + */ + +#include "unionfs.h" + +/* Delete all of the whiteouts in a given directory for rmdir. */ +int do_delete_whiteouts(struct dentry *dentry, int bindex, + struct unionfs_dir_state *namelist) +{ + int err = 0; + struct dentry *hidden_dir_dentry = NULL; + struct dentry *hidden_dentry; + char *name = NULL, *p; + struct inode *hidden_dir; + + int i; + struct list_head *pos; + struct filldir_node *cursor; + + /* Find out hidden parent dentry */ + hidden_dir_dentry = dtohd_index(dentry, bindex); + BUG_ON(!S_ISDIR(hidden_dir_dentry->d_inode->i_mode)); + hidden_dir = hidden_dir_dentry->d_inode; + BUG_ON(!S_ISDIR(hidden_dir->i_mode)); + + err = -ENOMEM; + name = __getname(); + if (!name) + goto out; + strcpy(name, WHPFX); + p = name + WHLEN; + + err = 0; + for (i = 0; !err && i < namelist->uds_size; i++) { + list_for_each(pos, &namelist->uds_list[i]) { + cursor = + list_entry(pos, struct filldir_node, file_list); + /* Only operate on whiteouts in this branch. */ + if (cursor->bindex != bindex) + continue; + if (!cursor->whiteout) + continue; + + strcpy(p, cursor->name); + hidden_dentry = + lookup_one_len(name, hidden_dir_dentry, + cursor->namelen + WHLEN); + if (IS_ERR(hidden_dentry)) { + err = PTR_ERR(hidden_dentry); + break; + } + if (hidden_dentry->d_inode) + err = vfs_unlink(hidden_dir, hidden_dentry, + NULL); + dput(hidden_dentry); + if (err) + break; + } + } + + __putname(name); + + /* After all of the removals, we should copy the attributes once. */ + fist_copy_attr_times(dentry->d_inode, hidden_dir_dentry->d_inode); + +out: + return err; +} + +/* Delete all of the whiteouts in a given directory for rmdir. */ +int delete_whiteouts(struct dentry *dentry, int bindex, + struct unionfs_dir_state *namelist) +{ + int err = 0; + struct dentry *hidden_dir_dentry = NULL; + struct super_block *sb; + char *name = NULL, *p; + struct inode *hidden_dir; + + struct sioq_args args; + + print_entry_location(); + + sb = dentry->d_sb; + unionfs_read_lock(sb); + + BUG_ON(!S_ISDIR(dentry->d_inode->i_mode)); + BUG_ON(bindex < dbstart(dentry)); + BUG_ON(bindex > dbend(dentry)); + err = is_robranch_super(sb, bindex); + if (err) + goto out; + + /* Find out hidden parent dentry */ + hidden_dir_dentry = dtohd_index(dentry, bindex); + BUG_ON(!S_ISDIR(hidden_dir_dentry->d_inode->i_mode)); + hidden_dir = hidden_dir_dentry->d_inode; + BUG_ON(!S_ISDIR(hidden_dir->i_mode)); + + err = -ENOMEM; + name = __getname(); + if (!name) + goto out; + strcpy(name, WHPFX); + p = name + WHLEN; + + err = 0; + mutex_lock(&hidden_dir->i_mutex); + + if (!inode_permission(hidden_dir, MAY_WRITE | MAY_EXEC)) + err = do_delete_whiteouts(dentry, bindex, namelist); + else { + args.deletewh.namelist = namelist; + args.deletewh.dentry = dentry; + args.deletewh.bindex = bindex; + run_sioq(__delete_whiteouts, &args); + err = args.err; + } + + mutex_unlock(&hidden_dir->i_mutex); + + out: + unionfs_read_unlock(sb); + print_exit_status(err); + return err; +} + +#define RD_NONE 0 +#define RD_CHECK_EMPTY 1 +/* The callback structure for check_empty. */ +struct unionfs_rdutil_callback { + int err; + int filldir_called; + struct unionfs_dir_state *rdstate; + int mode; +}; + +/* This filldir function makes sure only whiteouts exist within a directory. */ +static int readdir_util_callback(void *dirent, const char *name, int namelen, + loff_t offset, u64 ino, unsigned int d_type) +{ + int err = 0; + struct unionfs_rdutil_callback *buf = + (struct unionfs_rdutil_callback *)dirent; + int whiteout = 0; + struct filldir_node *found; + + print_entry_location(); + + buf->filldir_called = 1; + + if (name[0] == '.' + && (namelen == 1 || (name[1] == '.' && namelen == 2))) + goto out; + + if ((namelen > WHLEN) && !strncmp(name, WHPFX, WHLEN)) { + namelen -= WHLEN; + name += WHLEN; + whiteout = 1; + } + + found = find_filldir_node(buf->rdstate, name, namelen); + /* If it was found in the table there was a previous whiteout. */ + if (found) + goto out; + + /* If it wasn't found and isn't a whiteout, the directory isn't empty. */ + err = -ENOTEMPTY; + if ((buf->mode == RD_CHECK_EMPTY) && !whiteout) + goto out; + + err = add_filldir_node(buf->rdstate, name, namelen, + buf->rdstate->uds_bindex, whiteout); + + out: + buf->err = err; + print_exit_status(err); + return err; +} + +/* Is a directory logically empty? */ +int check_empty(struct dentry *dentry, struct unionfs_dir_state **namelist) +{ + int err = 0; + struct dentry *hidden_dentry = NULL; + struct super_block *sb; + struct file *hidden_file; + struct unionfs_rdutil_callback *buf = NULL; + int bindex, bstart, bend, bopaque; + + print_entry_location(); + + sb = dentry->d_sb; + + unionfs_read_lock(sb); + + BUG_ON(!S_ISDIR(dentry->d_inode->i_mode)); + + if ((err = unionfs_partial_lookup(dentry))) + goto out; + + bstart = dbstart(dentry); + bend = dbend(dentry); + bopaque = dbopaque(dentry); + if (0 <= bopaque && bopaque < bend) + bend = bopaque; + + buf = KMALLOC(sizeof(struct unionfs_rdutil_callback), GFP_KERNEL); + if (!buf) { + err = -ENOMEM; + goto out; + } + buf->err = 0; + buf->mode = RD_CHECK_EMPTY; + buf->rdstate = alloc_rdstate(dentry->d_inode, bstart); + if (!buf->rdstate) { + err = -ENOMEM; + goto out; + } + + /* Process the hidden directories with rdutil_callback as a filldir. */ + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + if (!hidden_dentry->d_inode) + continue; + if (!S_ISDIR(hidden_dentry->d_inode->i_mode)) + continue; + + DGET(hidden_dentry); + mntget(stohiddenmnt_index(sb, bindex)); + branchget(sb, bindex); + hidden_file = + DENTRY_OPEN(hidden_dentry, stohiddenmnt_index(sb, bindex), + O_RDONLY); + if (IS_ERR(hidden_file)) { + err = PTR_ERR(hidden_file); + DPUT(hidden_dentry); + branchput(sb, bindex); + goto out; + } + + do { + buf->filldir_called = 0; + buf->rdstate->uds_bindex = bindex; + err = vfs_readdir(hidden_file, + readdir_util_callback, buf); + if (buf->err) + err = buf->err; + } while ((err >= 0) && buf->filldir_called); + + /* fput calls dput for hidden_dentry */ + fput(hidden_file); + branchput(sb, bindex); + + if (err < 0) + goto out; + } + + out: + if (buf) { + if (namelist && !err) + *namelist = buf->rdstate; + else if (buf->rdstate) + free_rdstate(buf->rdstate); + KFREE(buf); + } + + unionfs_read_unlock(sb); + + print_exit_status(err); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/file.c @@ -0,0 +1,393 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: file.c,v 1.143 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +/* declarations for sparse */ +extern ssize_t unionfs_read(struct file *, char __user *, size_t, loff_t *); +extern ssize_t unionfs_write(struct file *, const char __user *, size_t, + loff_t *); + +/******************* + * File Operations * + *******************/ + +#ifndef UNIONFS_MMAP +/* SP: Disable unionfs_llseek, as use generic_file_llseek on upper file */ +static loff_t unionfs_llseek(struct file *file, loff_t offset, int origin) +{ + loff_t err; + struct file *hidden_file = NULL; + + print_entry_location(); + + dprint(PRINT_DEBUG, "unionfs_llseek: file=%p, offset=0x%llx, origin=%d\n", + file, offset, origin); + + if ((err = unionfs_file_revalidate(file, 0))) + goto out; + + hidden_file = ftohf(file); + /* always set hidden position to this one */ + hidden_file->f_pos = file->f_pos; + + memcpy(&(hidden_file->f_ra), &(file->f_ra), + sizeof(struct file_ra_state)); + + if (hidden_file->f_op && hidden_file->f_op->llseek) + err = hidden_file->f_op->llseek(hidden_file, offset, origin); + else + err = generic_file_llseek(hidden_file, offset, origin); + + if (err < 0) + goto out; + if (err != file->f_pos) { + file->f_pos = err; + // ION maybe this? + // file->f_pos = hidden_file->f_pos; + + file->f_version++; + } + out: + print_exit_status((int)err); + return err; +} +#endif +ssize_t __unionfs_read(struct file * file, char __user * buf, size_t count, + loff_t * ppos) +{ + int err = -EINVAL; + struct file *hidden_file = NULL; + loff_t pos = *ppos; + + print_file("entering __unionfs_read()", file); + + hidden_file = ftohf(file); + if (!hidden_file->f_op || !hidden_file->f_op->read) + goto out; + + err = hidden_file->f_op->read(hidden_file, buf, count, &pos); + *ppos = pos; + + out: + print_file("leaving __unionfs_read()", file); + + print_exit_status(err); + return err; +} + +ssize_t unionfs_read(struct file * file, char __user * buf, size_t count, + loff_t * ppos) +{ + int err = -EINVAL; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 0))) + goto out; + +#ifdef UNIONFS_MMAP + err = generic_file_read(file, buf, count, ppos); + if (err >= 0) + file_accessed(ftohf(file)); +#else + err = __unionfs_read(file, buf, count, ppos); +#endif + + out: + + print_exit_status(err); + return err; +} + +/* SP: Sendfile code not updated, but should be able to use + * generic_file_sendfile, as it would use readpage, which we now have */ +#ifdef SUPPORT_BROKEN_LOSETUP +static ssize_t unionfs_sendfile(struct file *file, loff_t * ppos, + size_t count, read_actor_t actor, void *target) +{ + ssize_t err; + struct file *hidden_file = NULL; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 0))) + goto out; + + hidden_file = ftohf(file); + + err = -EINVAL; + if (!hidden_file->f_op || !hidden_file->f_op->sendfile) + goto out; + + err = hidden_file->f_op->sendfile(hidden_file, ppos, count, actor, + target); + + out: + print_exit_status(err); + return err; +} +#endif +ssize_t __unionfs_write(struct file * file, const char __user * buf, + size_t count, loff_t * ppos) +{ + int err = -EINVAL; + struct file *hidden_file = NULL; + struct inode *inode; + struct inode *hidden_inode; + loff_t pos = *ppos; + int bstart, bend; + + print_entry_location(); + + inode = file->f_dentry->d_inode; + + bstart = fbstart(file); + bend = fbend(file); + + BUG_ON(bstart == -1); + + hidden_file = ftohf(file); + hidden_inode = hidden_file->f_dentry->d_inode; + + if (!hidden_file->f_op || !hidden_file->f_op->write) + goto out; + + /* adjust for append -- seek to the end of the file */ + if (file->f_flags & O_APPEND) + pos = inode->i_size; + + err = hidden_file->f_op->write(hidden_file, buf, count, &pos); + + /* + * copy ctime and mtime from lower layer attributes + * atime is unchanged for both layers + */ + if (err >= 0) + fist_copy_attr_times(inode, hidden_inode); + + *ppos = pos; + + /* update this inode's size */ + if (pos > inode->i_size) + inode->i_size = pos; + out: + print_exit_status(err); + return err; +} + +ssize_t unionfs_write(struct file * file, const char __user * buf, size_t count, + loff_t * ppos) +{ + int err = 0; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 1))) + goto out; +#ifdef UNIONFS_MMAP + err = generic_file_write(file, buf, count, ppos); +#else + err = __unionfs_write(file, buf, count, ppos); +#endif + out: + print_exit_status(err); + return err; +} + +static int unionfs_file_readdir(struct file *file, void *dirent, + filldir_t filldir) +{ + int err = -ENOTDIR; + print_entry_location(); + print_exit_status(err); + return err; +} + +static unsigned int unionfs_poll(struct file *file, poll_table * wait) +{ + unsigned int mask = DEFAULT_POLLMASK; + struct file *hidden_file = NULL; + + print_entry_location(); + + if (unionfs_file_revalidate(file, 0)) { + /* We should pretend an error happend. */ + mask = POLLERR | POLLIN | POLLOUT; + goto out; + } + + hidden_file = ftohf(file); + + if (!hidden_file->f_op || !hidden_file->f_op->poll) + goto out; + + mask = hidden_file->f_op->poll(hidden_file, wait); + + out: + print_exit_status(mask); + return mask; +} + +#ifndef UNIONFS_MMAP +static int __do_mmap(struct file *file, struct vm_area_struct *vma) +{ + int err; + struct file *hidden_file; + + print_entry_location(); + hidden_file = ftohf(file); + + err = -ENODEV; + if (!hidden_file->f_op || !hidden_file->f_op->mmap) + goto out; + + vma->vm_file = hidden_file; + err = hidden_file->f_op->mmap(hidden_file, vma); + get_file(hidden_file); /* make sure it doesn't get freed on us */ + fput(file); /* no need to keep extra ref on ours */ + out: + print_exit_status(err); + return err; +} +#endif +/* SP: mmap code now maps upper file + * like old code, will only copyup at this point, it's possible to copyup + * in writepage(), but I haven't bothered with that, as only apt-get seem + * to want to write to a shared/write mapping + */ +static int unionfs_mmap(struct file *file, struct vm_area_struct *vma) +{ + int err = 0; + int willwrite; + + print_entry_location(); + + /* This might could be deferred to mmap's writepage. */ + willwrite = ((vma->vm_flags | VM_SHARED | VM_WRITE) == vma->vm_flags); + if ((err = unionfs_file_revalidate(file, willwrite))) + goto out; +#ifdef UNIONFS_MMAP + err = generic_file_mmap(file, vma); + if (err) { + printk("unionfs_mmap: generic_file_mmap failed\n"); + } +#else + err = __do_mmap(file, vma); +#endif + + out: + print_exit_status(err); + return err; +} + +/* SP: disabled as use the generic file_fsync */ +#ifndef UNIONFS_MMAP +static int unionfs_fsync(struct file *file, struct dentry *dentry, int datasync) +{ + int err; + struct file *hidden_file = NULL; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 1))) + goto out; + + hidden_file = ftohf(file); + + err = -EINVAL; + if (!hidden_file->f_op || !hidden_file->f_op->fsync) + goto out; + + mutex_lock(&hidden_file->f_dentry->d_inode->i_mutex); + err = hidden_file->f_op->fsync(hidden_file, hidden_file->f_dentry, + datasync); + mutex_unlock(&hidden_file->f_dentry->d_inode->i_mutex); + + out: + print_exit_status(err); + return err; +} +#endif + +/* SP: disabled as none of the other in kernel fs's seem to use it */ +static int unionfs_fasync(int fd, struct file *file, int flag) +{ + int err = 0; + struct file *hidden_file = NULL; + + print_entry_location(); + + if ((err = unionfs_file_revalidate(file, 1))) + goto out; + + hidden_file = ftohf(file); + + if (hidden_file->f_op && hidden_file->f_op->fasync) + err = hidden_file->f_op->fasync(fd, hidden_file, flag); + + out: + print_exit_status(err); + return err; +} + +struct file_operations unionfs_main_fops = { +#ifdef UNIONFS_MMAP + .llseek = generic_file_llseek, +#else + .llseek = unionfs_llseek, +#endif + .read = unionfs_read, + .write = unionfs_write, + .readdir = unionfs_file_readdir, + .poll = unionfs_poll, + .unlocked_ioctl = unionfs_ioctl, + .mmap = unionfs_mmap, + .open = unionfs_open, + .flush = unionfs_flush, + .release = unionfs_file_release, +#ifdef UNIONFS_MMAP + .fsync = file_fsync, +#else + .fsync = unionfs_fsync, +#endif + .fasync = unionfs_fasync, +#ifdef SUPPORT_BROKEN_LOSETUP + .sendfile = unionfs_sendfile, +#endif +}; + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/inode.c @@ -0,0 +1,1034 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: inode.c,v 1.275 2006/10/31 18:05:33 yiannos Exp $ + */ + +#include "unionfs.h" + +/* declarations added for "sparse" */ +extern struct dentry *unionfs_lookup(struct inode *, struct dentry *, + struct nameidata *); +extern int unionfs_readlink(struct dentry *dentry, char __user * buf, + int bufsiz); +extern void unionfs_put_link(struct dentry *dentry, struct nameidata *nd, + void *cookie); + +static int unionfs_create(struct inode *parent, struct dentry *dentry, + int mode, struct nameidata *nd) +{ + int err = 0; + struct dentry *hidden_dentry = NULL; + struct dentry *whiteout_dentry = NULL; + struct dentry *new_hidden_dentry; + struct dentry *hidden_parent_dentry = NULL; + int bindex = 0, bstart; + char *name = NULL; + + print_entry_location(); + lock_dentry(dentry); + print_dentry("IN unionfs_create", dentry); + + /* We start out in the leftmost branch. */ + bstart = dbstart(dentry); + hidden_dentry = dtohd(dentry); + + /* check if whiteout exists in this branch, i.e. lookup .wh.foo first */ + name = alloc_whname(dentry->d_name.name, dentry->d_name.len); + if (IS_ERR(name)) { + err = PTR_ERR(name); + goto out; + } + + whiteout_dentry = + LOOKUP_ONE_LEN(name, hidden_dentry->d_parent, + dentry->d_name.len + WHLEN); + if (IS_ERR(whiteout_dentry)) { + err = PTR_ERR(whiteout_dentry); + whiteout_dentry = NULL; + goto out; + } + + if (whiteout_dentry->d_inode) { + /* .wh.foo has been found. */ + /* First truncate it and then rename it to foo (hence having + * the same overall effect as a normal create. + * + * XXX: This is not strictly correct. If we have unlinked the + * file and it still has a reference count, then we should + * actually unlink the whiteout so that user's data isn't + * hosed over. + */ + struct dentry *hidden_dir_dentry; + struct iattr newattrs; + + mutex_lock(&whiteout_dentry->d_inode->i_mutex); + newattrs.ia_valid = ATTR_CTIME | ATTR_ATIME + | ATTR_MTIME | ATTR_UID | ATTR_GID | ATTR_FORCE + | ATTR_KILL_SUID | ATTR_KILL_SGID; + + newattrs.ia_mode = mode & ~current->fs->umask; + newattrs.ia_uid = current->fsuid; + newattrs.ia_gid = current->fsgid; + + if (whiteout_dentry->d_inode->i_size != 0) { + newattrs.ia_valid |= ATTR_SIZE; + newattrs.ia_size = 0; + } + + err = notify_change(whiteout_dentry, NULL, &newattrs); + + mutex_unlock(&whiteout_dentry->d_inode->i_mutex); + + if (err) + printk(KERN_WARNING + "unionfs: %s:%d: notify_change failed: %d, ignoring..\n", + __FILE__, __LINE__, err); + + new_hidden_dentry = dtohd(dentry); + DGET(new_hidden_dentry); + + hidden_dir_dentry = GET_PARENT(whiteout_dentry); + lock_rename(hidden_dir_dentry, hidden_dir_dentry); + + if (!(err = is_robranch_super(dentry->d_sb, bstart))) { + err = + vfs_rename(hidden_dir_dentry->d_inode, + whiteout_dentry, NULL, + hidden_dir_dentry->d_inode, + new_hidden_dentry, NULL); + } + if (!err) { + fist_copy_attr_timesizes(parent, + new_hidden_dentry->d_parent-> + d_inode); + parent->i_nlink = get_nlinks(parent); + } + + unlock_rename(hidden_dir_dentry, hidden_dir_dentry); + DPUT(hidden_dir_dentry); + + DPUT(new_hidden_dentry); + + if (err) { + /* exit if the error returned was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + goto out; + /* We were not able to create the file in this branch, + * so, we try to create it in one branch to left + */ + bstart--; + } else { + /* reset the unionfs dentry to point to the .wh.foo entry. */ + + /* Discard any old reference. */ + DPUT(dtohd(dentry)); + + /* Trade one reference to another. */ + set_dtohd_index(dentry, bstart, whiteout_dentry); + whiteout_dentry = NULL; + + err = unionfs_interpose(dentry, parent->i_sb, 0); + goto out; + } + } + + for (bindex = bstart; bindex >= 0; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) { + /* if hidden_dentry is NULL, create the entire + * dentry directory structure in branch 'bindex'. + * hidden_dentry will NOT be null when bindex == bstart + * because lookup passed as a negative unionfs dentry + * pointing to a lone negative underlying dentry */ + hidden_dentry = create_parents(parent, dentry, bindex); + if (!hidden_dentry || IS_ERR(hidden_dentry)) { + if (IS_ERR(hidden_dentry)) + err = PTR_ERR(hidden_dentry); + continue; + } + } + + checkinode(parent, "unionfs_create"); + + hidden_parent_dentry = lock_parent(hidden_dentry); + if (IS_ERR(hidden_parent_dentry)) { + err = PTR_ERR(hidden_parent_dentry); + goto out; + } + /* We shouldn't create things in a read-only branch. */ + if (!(err = is_robranch_super(dentry->d_sb, bindex))) { + //DQ: vfs_create has a different prototype in 2.6 + err = vfs_create(hidden_parent_dentry->d_inode, + hidden_dentry, mode, nd); + } + if (err || !hidden_dentry->d_inode) { + unlock_dir(hidden_parent_dentry); + + /* break out of for loop if the error wasn't -EROFS */ + if (!IS_COPYUP_ERR(err)) + break; + } else { + err = unionfs_interpose(dentry, parent->i_sb, 0); + if (!err) { + fist_copy_attr_timesizes(parent, + hidden_parent_dentry-> + d_inode); + /* update number of links on parent directory */ + parent->i_nlink = get_nlinks(parent); + } + unlock_dir(hidden_parent_dentry); + break; + } + } + + out: + DPUT(whiteout_dentry); + KFREE(name); + + print_dentry("OUT unionfs_create :", dentry); + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +struct dentry *unionfs_lookup(struct inode *parent, struct dentry *dentry, + struct nameidata *nd) +{ + struct nameidata lowernd; + + if(nd) + memcpy(&lowernd, nd, sizeof(struct nameidata)); + else + memset(&lowernd, 0, sizeof(struct nameidata)); + + /* The locking is done by unionfs_lookup_backend. */ + return unionfs_lookup_backend(dentry, &lowernd, INTERPOSE_LOOKUP); +} + +static int unionfs_link(struct dentry *old_dentry, struct inode *dir, + struct dentry *new_dentry) +{ + int err = 0; + struct dentry *hidden_old_dentry = NULL; + struct dentry *hidden_new_dentry = NULL; + struct dentry *hidden_dir_dentry = NULL; + struct dentry *whiteout_dentry; + char *name = NULL; + + print_entry_location(); + double_lock_dentry(new_dentry, old_dentry); + + hidden_new_dentry = dtohd(new_dentry); + + /* check if whiteout exists in the branch of new dentry, i.e. lookup + * .wh.foo first. If present, delete it */ + name = alloc_whname(new_dentry->d_name.name, new_dentry->d_name.len); + if (IS_ERR(name)) { + err = PTR_ERR(name); + goto out; + } + + whiteout_dentry = + LOOKUP_ONE_LEN(name, hidden_new_dentry->d_parent, + new_dentry->d_name.len + WHLEN); + if (IS_ERR(whiteout_dentry)) { + err = PTR_ERR(whiteout_dentry); + goto out; + } + + if (!whiteout_dentry->d_inode) { + DPUT(whiteout_dentry); + whiteout_dentry = NULL; + } else { + /* found a .wh.foo entry, unlink it and then call vfs_link() */ + hidden_dir_dentry = lock_parent(whiteout_dentry); + if (! + (err = + is_robranch_super(new_dentry->d_sb, + dbstart(new_dentry)))) { + err = + vfs_unlink(hidden_dir_dentry->d_inode, + whiteout_dentry, NULL); + } + fist_copy_attr_times(dir, hidden_dir_dentry->d_inode); + dir->i_nlink = get_nlinks(dir); + unlock_dir(hidden_dir_dentry); + hidden_dir_dentry = NULL; + DPUT(whiteout_dentry); + if (err) + goto out; + } + + if (dbstart(old_dentry) != dbstart(new_dentry)) { + hidden_new_dentry = + create_parents(dir, new_dentry, dbstart(old_dentry)); + err = PTR_ERR(hidden_new_dentry); + if (IS_COPYUP_ERR(err)) + goto docopyup; + if (!hidden_new_dentry || IS_ERR(hidden_new_dentry)) + goto out; + } + hidden_new_dentry = dtohd(new_dentry); + hidden_old_dentry = dtohd(old_dentry); + + BUG_ON(dbstart(old_dentry) != dbstart(new_dentry)); + hidden_dir_dentry = lock_parent(hidden_new_dentry); + if (!(err = is_robranch(old_dentry))) + err = + vfs_link(hidden_old_dentry, NULL, + hidden_dir_dentry->d_inode, + hidden_new_dentry, NULL); + unlock_dir(hidden_dir_dentry); + + docopyup: + if (IS_COPYUP_ERR(err)) { + int old_bstart = dbstart(old_dentry); + int bindex; + + for (bindex = old_bstart - 1; bindex >= 0; bindex--) { + err = + copyup_dentry(old_dentry->d_parent-> + d_inode, old_dentry, + old_bstart, bindex, NULL, + old_dentry->d_inode->i_size); + if (!err) { + hidden_new_dentry = + create_parents(dir, new_dentry, bindex); + hidden_old_dentry = dtohd(old_dentry); + hidden_dir_dentry = + lock_parent(hidden_new_dentry); + /* do vfs_link */ + err = vfs_link(hidden_old_dentry, NULL, + hidden_dir_dentry->d_inode, + hidden_new_dentry, NULL); + unlock_dir(hidden_dir_dentry); + goto check_link; + } + } + goto out; + } + check_link: + if (err || !hidden_new_dentry->d_inode) + goto out; + + /* Its a hard link, so use the same inode */ + new_dentry->d_inode = IGRAB(old_dentry->d_inode); + d_instantiate(new_dentry, new_dentry->d_inode); + fist_copy_attr_all(dir, hidden_new_dentry->d_parent->d_inode); + /* propagate number of hard-links */ + old_dentry->d_inode->i_nlink = get_nlinks(old_dentry->d_inode); + + out: + if (!new_dentry->d_inode) + d_drop(new_dentry); + + KFREE(name); + + unlock_dentry(new_dentry); + unlock_dentry(old_dentry); + + print_exit_status(err); + return err; +} + +static int unionfs_symlink(struct inode *dir, struct dentry *dentry, + const char *symname) +{ + int err = 0; + struct dentry *hidden_dentry = NULL; + struct dentry *whiteout_dentry = NULL; + struct dentry *hidden_dir_dentry = NULL; + umode_t mode; + int bindex = 0, bstart; + char *name = NULL; + + print_entry_location(); + lock_dentry(dentry); + print_dentry("IN unionfs_symlink", dentry); + + /* We start out in the leftmost branch. */ + bstart = dbstart(dentry); + + hidden_dentry = dtohd(dentry); + + /* check if whiteout exists in this branch, i.e. lookup .wh.foo first. If present, delete it */ + name = alloc_whname(dentry->d_name.name, dentry->d_name.len); + if (IS_ERR(name)) { + err = PTR_ERR(name); + goto out; + } + + whiteout_dentry = + LOOKUP_ONE_LEN(name, hidden_dentry->d_parent, + dentry->d_name.len + WHLEN); + if (IS_ERR(whiteout_dentry)) { + err = PTR_ERR(whiteout_dentry); + goto out; + } + + if (!whiteout_dentry->d_inode) { + DPUT(whiteout_dentry); + whiteout_dentry = NULL; + } else { + /* found a .wh.foo entry, unlink it and then call vfs_symlink() */ + hidden_dir_dentry = lock_parent(whiteout_dentry); + + print_dentry("HDD", hidden_dir_dentry); + print_dentry("WD", whiteout_dentry); + + if (!(err = is_robranch_super(dentry->d_sb, bstart))) { + err = + vfs_unlink(hidden_dir_dentry->d_inode, + whiteout_dentry, NULL); + } + DPUT(whiteout_dentry); + + fist_copy_attr_times(dir, hidden_dir_dentry->d_inode); + /* propagate number of hard-links */ + dir->i_nlink = get_nlinks(dir); + + unlock_dir(hidden_dir_dentry); + + if (err) { + /* exit if the error returned was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + goto out; + /* should now try to create symlink in the another branch */ + bstart--; + } + } + + /* deleted whiteout if it was present, now do a normal vfs_symlink() with + possible recursive directory creation */ + for (bindex = bstart; bindex >= 0; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) { + /* if hidden_dentry is NULL, create the entire + * dentry directory structure in branch 'bindex'. hidden_dentry will NOT be null when + * bindex == bstart because lookup passed as a negative unionfs dentry pointing to a + * lone negative underlying dentry */ + hidden_dentry = create_parents(dir, dentry, bindex); + if (!hidden_dentry || IS_ERR(hidden_dentry)) { + if (IS_ERR(hidden_dentry)) { + err = PTR_ERR(hidden_dentry); + } + dprint(PRINT_DEBUG, + "hidden dentry NULL (or error) for bindex = %d\n", + bindex); + continue; + } + } + + hidden_dir_dentry = lock_parent(hidden_dentry); + + if (!(err = is_robranch_super(dentry->d_sb, bindex))) { + mode = S_IALLUGO; + err = + vfs_symlink(hidden_dir_dentry->d_inode, + hidden_dentry, NULL, symname); + } + unlock_dir(hidden_dir_dentry); + + if (err || !hidden_dentry->d_inode) { + /* break out of for loop if error returned was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + break; + } else { + err = unionfs_interpose(dentry, dir->i_sb, 0); + if (!err) { + fist_copy_attr_timesizes(dir, + hidden_dir_dentry-> + d_inode); + /* update number of links on parent directory */ + dir->i_nlink = get_nlinks(dir); + } + break; + } + } + + out: + if (!dentry->d_inode) + d_drop(dentry); + + KFREE(name); + print_dentry("OUT unionfs_symlink :", dentry); + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +static int unionfs_mkdir(struct inode *parent, struct dentry *dentry, int mode) +{ + int err = 0; + struct dentry *hidden_dentry = NULL, *whiteout_dentry = NULL; + struct dentry *hidden_parent_dentry = NULL; + int bindex = 0, bstart; + char *name = NULL; + int whiteout_unlinked = 0; + struct sioq_args args; + + print_entry_location(); + lock_dentry(dentry); + print_dentry("IN unionfs_mkdir", dentry); + bstart = dbstart(dentry); + + hidden_dentry = dtohd(dentry); + + // check if whiteout exists in this branch, i.e. lookup .wh.foo first + name = alloc_whname(dentry->d_name.name, dentry->d_name.len); + if (IS_ERR(name)) { + err = PTR_ERR(name); + goto out; + } + + whiteout_dentry = + LOOKUP_ONE_LEN(name, hidden_dentry->d_parent, + dentry->d_name.len + WHLEN); + if (IS_ERR(whiteout_dentry)) { + err = PTR_ERR(whiteout_dentry); + goto out; + } + + if (!whiteout_dentry->d_inode) { + DPUT(whiteout_dentry); + whiteout_dentry = NULL; + } else { + hidden_parent_dentry = lock_parent(whiteout_dentry); + + //found a.wh.foo entry, remove it then do vfs_mkdir + if (!(err = is_robranch_super(dentry->d_sb, bstart))) { + args.unlink.parent = hidden_parent_dentry->d_inode; + args.unlink.dentry = whiteout_dentry; + run_sioq(__unionfs_unlink, &args); + err = args.err; + } + DPUT(whiteout_dentry); + + unlock_dir(hidden_parent_dentry); + + if (err) { + /* exit if the error returned was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + goto out; + bstart--; + } else { + whiteout_unlinked = 1; + } + } + + for (bindex = bstart; bindex >= 0; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) { + hidden_dentry = create_parents(parent, dentry, bindex); + if (!hidden_dentry || IS_ERR(hidden_dentry)) { + dprint(PRINT_DEBUG, + "hidden dentry NULL for bindex = %d\n", + bindex); + continue; + } + } + + hidden_parent_dentry = lock_parent(hidden_dentry); + if (IS_ERR(hidden_parent_dentry)) { + err = PTR_ERR(hidden_parent_dentry); + goto out; + } + if (!(err = is_robranch_super(dentry->d_sb, bindex))) { + err = + vfs_mkdir(hidden_parent_dentry->d_inode, + hidden_dentry, NULL, mode); + } + unlock_dir(hidden_parent_dentry); + + /* XXX this could potentially return a negative hidden_dentry! */ + if (err || !hidden_dentry->d_inode) { + /* break out of for loop if error returned was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + break; + } else { + int i; + int bend = dbend(dentry); + + for (i = bindex + 1; i < bend; i++) { + if (dtohd_index(dentry, i)) { + DPUT(dtohd_index(dentry, i)); + set_dtohd_index(dentry, i, NULL); + } + } + bend = bindex; + set_dbend(dentry, bend); + + err = unionfs_interpose(dentry, parent->i_sb, 0); + if (!err) { + fist_copy_attr_timesizes(parent, + hidden_parent_dentry-> + d_inode); + /* update number of links on parent directory */ + parent->i_nlink = get_nlinks(parent); + } + + err = make_dir_opaque(dentry, dbstart(dentry)); + if (err) { + dprint(PRINT_DEBUG, + "mkdir: error creating directory override entry: %d\n", + err); + goto out; + } + break; + } + } + + out: + if (!dentry->d_inode) + d_drop(dentry); + + KFREE(name); + + print_dentry("OUT unionfs_mkdir :", dentry); + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +static int unionfs_mknod(struct inode *dir, struct dentry *dentry, int mode, + dev_t dev) +{ + int err = 0; + struct dentry *hidden_dentry = NULL, *whiteout_dentry = NULL; + struct dentry *hidden_parent_dentry = NULL; + int bindex = 0, bstart; + char *name = NULL; + int whiteout_unlinked = 0; + + print_entry_location(); + lock_dentry(dentry); + print_dentry("IN unionfs_mknod", dentry); + bstart = dbstart(dentry); + + hidden_dentry = dtohd(dentry); + + // check if whiteout exists in this branch, i.e. lookup .wh.foo first + name = alloc_whname(dentry->d_name.name, dentry->d_name.len); + if (IS_ERR(name)) { + err = PTR_ERR(name); + goto out; + } + + whiteout_dentry = + LOOKUP_ONE_LEN(name, hidden_dentry->d_parent, + dentry->d_name.len + WHLEN); + if (IS_ERR(whiteout_dentry)) { + err = PTR_ERR(whiteout_dentry); + goto out; + } + + if (!whiteout_dentry->d_inode) { + DPUT(whiteout_dentry); + whiteout_dentry = NULL; + } else { + /* found .wh.foo, unlink it */ + hidden_parent_dentry = lock_parent(whiteout_dentry); + + //found a.wh.foo entry, remove it then do vfs_mkdir + if (!(err = is_robranch_super(dentry->d_sb, bstart))) + err = vfs_unlink(hidden_parent_dentry->d_inode, + whiteout_dentry, NULL); + DPUT(whiteout_dentry); + + unlock_dir(hidden_parent_dentry); + + if (err) { + if (!IS_COPYUP_ERR(err)) + goto out; + + bstart--; + } else { + whiteout_unlinked = 1; + } + } + + for (bindex = bstart; bindex >= 0; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) { + hidden_dentry = create_parents(dir, dentry, bindex); + if (!hidden_dentry || IS_ERR(hidden_dentry)) { + dprint(PRINT_DEBUG, + "hidden dentry NULL for bindex = %d\n", + bindex); + continue; + } + } + + hidden_parent_dentry = lock_parent(hidden_dentry); + if (IS_ERR(hidden_parent_dentry)) { + err = PTR_ERR(hidden_parent_dentry); + goto out; + } + if (!(err = is_robranch_super(dentry->d_sb, bindex))) { + err = vfs_mknod(hidden_parent_dentry->d_inode, + hidden_dentry, NULL, mode, dev); + } + /* XXX this could potentially return a negative hidden_dentry! */ + if (err || !hidden_dentry->d_inode) { + unlock_dir(hidden_parent_dentry); + /* break out of for, if error was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + break; + } else { + err = unionfs_interpose(dentry, dir->i_sb, 0); + if (!err) { + fist_copy_attr_timesizes(dir, + hidden_parent_dentry-> + d_inode); + /* update number of links on parent directory */ + dir->i_nlink = get_nlinks(dir); + } + unlock_dir(hidden_parent_dentry); + + break; + } + } + + out: + if (!dentry->d_inode) + d_drop(dentry); + + if (name) { + KFREE(name); + } + + print_dentry("OUT unionfs_mknod :", dentry); + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +int unionfs_readlink(struct dentry *dentry, char __user * buf, int bufsiz) +{ + int err; + struct dentry *hidden_dentry; + + print_entry_location(); + lock_dentry(dentry); + hidden_dentry = dtohd(dentry); + print_dentry("unionfs_readlink IN", dentry); + + if (!hidden_dentry->d_inode->i_op || + !hidden_dentry->d_inode->i_op->readlink) { + err = -EINVAL; + goto out; + } + + err = hidden_dentry->d_inode->i_op->readlink(hidden_dentry, + buf, bufsiz); + if (err > 0) + fist_copy_attr_atime(dentry->d_inode, hidden_dentry->d_inode); + + out: + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +/* We don't lock the dentry here, because readlink does the heavy lifting. */ +static void *unionfs_follow_link(struct dentry *dentry, struct nameidata *nd) +{ + char *buf; + int len = PAGE_SIZE, err; + mm_segment_t old_fs; + + print_entry_location(); + + /* This is freed by the put_link method assuming a successful call. */ + buf = (char *)KMALLOC(len, GFP_KERNEL); + if (!buf) { + err = -ENOMEM; + goto out; + } + + /* read the symlink, and then we will follow it */ + old_fs = get_fs(); + set_fs(KERNEL_DS); + err = dentry->d_inode->i_op->readlink(dentry, (char __user *)buf, len); + set_fs(old_fs); + if (err < 0) { + KFREE(buf); + buf = NULL; + goto out; + } + buf[err] = 0; + nd_set_link(nd, buf); + err = 0; + + out: + print_exit_status(err); + return ERR_PTR(err); +} + +void unionfs_put_link(struct dentry *dentry, struct nameidata *nd, void *cookie) +{ + char *link; + print_entry_location(); + link = nd_get_link(nd); + KFREE(link); + print_exit_location(); +} + +/* Basically copied from the kernel vfs permission(), but we've changed + * the following: (1) the IS_RDONLY check is skipped, and (2) if you set + * the mount option `nfsperms=insceure', we assume that -EACCES means that + * the export is read-only and we should check standard Unix permissions. + * This means that NFS ACL checks (or other advanced permission features) + * are bypassed. + */ +static int unionfs_inode_permission(struct inode *inode, int mask, struct nameidata *nd, + int bindex) +{ + int retval, submask; + + if (mask & MAY_WRITE) { + /* The first branch is allowed to be really readonly. */ + if (bindex == 0) { + umode_t mode = inode->i_mode; + if (IS_RDONLY(inode) && (S_ISREG(mode) || S_ISDIR(mode) + || S_ISLNK(mode))) + return -EROFS; + } + /* + * Nobody gets write access to an immutable file. + */ + if (IS_IMMUTABLE(inode)) + return -EACCES; + } + + /* Ordinary permission routines do not understand MAY_APPEND. */ + submask = mask & ~MAY_APPEND; + if (inode->i_op && inode->i_op->permission) { + retval = inode->i_op->permission(inode, submask); + if ((retval == -EACCES) && (submask & MAY_WRITE) && + (!strcmp("nfs", (inode)->i_sb->s_type->name)) && + (nd) && (nd->path.mnt) && (nd->path.mnt->mnt_sb) && + (branchperms(nd->path.mnt->mnt_sb, bindex) & MAY_NFSRO)) { + retval = generic_permission(inode, submask, NULL); + } + } else { + retval = generic_permission(inode, submask, NULL); + } + + if (retval && retval != -EROFS) /* ignore EROFS */ + return retval; + + /* + * skip the LSM permission check. This means unionfs will wrongly + * copy up a LSM non-writable/non-readable file on a readonly branch + * to a read-write branch leading to odd behaviour. Until the mess + * of the LSM interface changes are resolved, there's nothing else + * that can be done. + * retval = security_inode_permission(inode, mask, nd); + */ + return ((retval == -EROFS) ? 0 : retval); /* ignore EROFS */ +} + +static int unionfs_permission(struct inode *inode, int mask) +{ + struct inode *hidden_inode = NULL; + int err = 0; + int bindex, bstart, bend; + const int is_file = !S_ISDIR(inode->i_mode); + const int write_mask = (mask & MAY_WRITE) && !(mask & MAY_READ); + + print_entry_location(); + + bstart = ibstart(inode); + bend = ibend(inode); + + print_inode("IN unionfs_permission", inode); + + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_inode = itohi_index(inode, bindex); + if (!hidden_inode) + continue; + + /* check the condition for D-F-D underlying files/directories, + * we dont have to check for files, if we are checking for + * directories. + */ + if (!is_file && !S_ISDIR(hidden_inode->i_mode)) + continue; + /* We use our own special version of permission, such that + * only the first branch returns -EROFS. */ + err = unionfs_inode_permission(hidden_inode, mask, NULL, bindex); + /* The permissions are an intersection of the overall directory + * permissions, so we fail if one fails. */ + if (err) + goto out; + /* only the leftmost file matters. */ + if (is_file || write_mask) { + if (is_file && write_mask) { + err = get_write_access(hidden_inode); + if (!err) + put_write_access(hidden_inode); + } + break; + } + } + + out: + print_exit_status(err); + return err; +} + +static int unionfs_setattr(struct dentry *dentry, struct iattr *ia) +{ + int err = 0; + struct dentry *hidden_dentry; + struct inode *inode = NULL; + struct inode *hidden_inode = NULL; + int bstart, bend, bindex; + int i; + int copyup = 0; + + print_entry_location(); + lock_dentry(dentry); + bstart = dbstart(dentry); + bend = dbend(dentry); + inode = dentry->d_inode; + + for (bindex = bstart; (bindex <= bend) || (bindex == bstart); bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + BUG_ON(hidden_dentry->d_inode == NULL); + + /* If the file is on a read only branch */ + if (is_robranch_super(dentry->d_sb, bindex) + || IS_RDONLY(hidden_dentry->d_inode)) { + if (copyup || (bindex != bstart)) + continue; + /* Only if its the leftmost file, copyup the file */ + for (i = bstart - 1; i >= 0; i--) { + loff_t size = dentry->d_inode->i_size; + if (ia->ia_valid & ATTR_SIZE) + size = ia->ia_size; + err = copyup_dentry(dentry->d_parent->d_inode, + dentry, bstart, i, NULL, + size); + + if (!err) { + copyup = 1; + hidden_dentry = dtohd(dentry); + break; + } + /* if error is in the leftmost f/s, pass it up */ + if (i == 0) + goto out; + } + + } + /* + * mode change is for clearing setuid/setgid bits. Allow lower fs + * to interpret this in its own way. + */ + if (ia->ia_valid & (ATTR_KILL_SUID | ATTR_KILL_SGID)) + ia->ia_valid &= ~ATTR_MODE; + + err = notify_change(hidden_dentry, NULL, ia); + if (err) + goto out; + break; + } +#ifdef UNIONFS_MMAP + /* + * SP: notify_change will change the lower file's size, + * but we need to truncate the page tables, so need to call + * vmtruncate() + */ + + if (ia->ia_valid & ATTR_SIZE) { + if (ia->ia_size != i_size_read(inode)) { + err = vmtruncate(inode, ia->ia_size); + if (err) { + printk("unionfs_setattr: vmtruncate failed\n"); + } + } + } +#endif + /* get the size from the first hidden inode */ + hidden_inode = itohi(dentry->d_inode); + checkinode(inode, "unionfs_setattr"); + fist_copy_attr_all(inode, hidden_inode); + + out: + unlock_dentry(dentry); + checkinode(inode, "post unionfs_setattr"); + print_exit_status(err); + return err; +} + +struct inode_operations unionfs_symlink_iops = { + .readlink = unionfs_readlink, + .permission = unionfs_permission, + .follow_link = unionfs_follow_link, + .setattr = unionfs_setattr, + .put_link = unionfs_put_link, +}; + +struct inode_operations unionfs_dir_iops = { + .create = unionfs_create, + .lookup = unionfs_lookup, + .link = unionfs_link, + .unlink = unionfs_unlink, + .symlink = unionfs_symlink, + .mkdir = unionfs_mkdir, + .rmdir = unionfs_rmdir, + .mknod = unionfs_mknod, + .rename = unionfs_rename, + .permission = unionfs_permission, + .setattr = unionfs_setattr, + .setxattr = unionfs_setxattr, + .getxattr = unionfs_getxattr, + .removexattr = unionfs_removexattr, + .listxattr = unionfs_listxattr, +}; + +struct inode_operations unionfs_main_iops = { + .permission = unionfs_permission, + .setattr = unionfs_setattr, + .setxattr = unionfs_setxattr, + .getxattr = unionfs_getxattr, + .removexattr = unionfs_removexattr, + .listxattr = unionfs_listxattr, +}; + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/lookup.c @@ -0,0 +1,524 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: lookup.c,v 1.49 2006/10/31 18:05:33 yiannos Exp $ + */ + +#include "unionfs.h" + +static int is_opaque_dir(struct dentry *dentry, int bindex); +static int is_validname(const char *name); + +struct dentry *unionfs_lookup_backend(struct dentry *dentry, struct nameidata *nd, int lookupmode) +{ + int err = 0; + struct dentry *hidden_dentry = NULL; + struct dentry *wh_hidden_dentry = NULL; + struct dentry *hidden_dir_dentry = NULL; + struct dentry *parent_dentry = NULL; + int bindex, bstart, bend, bopaque; + int dentry_count = 0; /* Number of positive dentries. */ + int first_dentry_offset = -1; + struct dentry *first_hidden_dentry = NULL; + int locked_parent = 0; + int locked_child = 0; + + int opaque; + char *whname = NULL; + const char *name; + int namelen; + + print_entry("mode = %d", lookupmode); + + /* We should already have a lock on this dentry in the case of a + * partial lookup, or a revalidation. Otherwise it is returned from + * new_dentry_private_data already locked. */ + if (lookupmode == INTERPOSE_PARTIAL || lookupmode == INTERPOSE_REVAL + || lookupmode == INTERPOSE_REVAL_NEG) { + verify_locked(dentry); + } else { + BUG_ON(dtopd_nocheck(dentry) != NULL); + locked_child = 1; + } + if (lookupmode != INTERPOSE_PARTIAL) + if ((err = new_dentry_private_data(dentry))) + goto out; + /* must initialize dentry operations */ + dentry->d_op = &unionfs_dops; + + parent_dentry = GET_PARENT(dentry); + /* We never partial lookup the root directory. */ + if (parent_dentry != dentry) { + lock_dentry(parent_dentry); + locked_parent = 1; + } else { + DPUT(parent_dentry); + parent_dentry = NULL; + goto out; + } + + print_dentry("IN unionfs_lookup (parent)", parent_dentry); + print_dentry("IN unionfs_lookup (child)", dentry); + + name = dentry->d_name.name; + namelen = dentry->d_name.len; + + /* No dentries should get created for possible whiteout names. */ + if (!is_validname(name)) { + err = -EPERM; + goto out_free; + } + + /* Now start the actual lookup procedure. */ + bstart = dbstart(parent_dentry); + bend = dbend(parent_dentry); + bopaque = dbopaque(parent_dentry); + BUG_ON(bstart < 0); + + /* It would be ideal if we could convert partial lookups to only have + * to do this work when they really need to. It could probably improve + * performance quite a bit, and maybe simplify the rest of the code. */ + if (lookupmode == INTERPOSE_PARTIAL) { + bstart++; + if ((bopaque != -1) && (bopaque < bend)) + bend = bopaque; + } + + dprint(PRINT_DEBUG, "bstart = %d, bend = %d\n", bstart, bend); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (lookupmode == INTERPOSE_PARTIAL && hidden_dentry) + continue; + BUG_ON(hidden_dentry != NULL); + + hidden_dir_dentry = dtohd_index(parent_dentry, bindex); + + /* if the parent hidden dentry does not exist skip this */ + if (!(hidden_dir_dentry && hidden_dir_dentry->d_inode)) + continue; + + /* also skip it if the parent isn't a directory. */ + if (!S_ISDIR(hidden_dir_dentry->d_inode->i_mode)) + continue; + + /* Reuse the whiteout name because its value doesn't change. */ + if (!whname) { + whname = alloc_whname(name, namelen); + if (IS_ERR(whname)) { + err = PTR_ERR(whname); + goto out_free; + } + } + + /* check if whiteout exists in this branch: lookup .wh.foo */ + wh_hidden_dentry = LOOKUP_ONE_LEN(whname, hidden_dir_dentry, + namelen + WHLEN); + if (IS_ERR(wh_hidden_dentry)) { + DPUT(first_hidden_dentry); + err = PTR_ERR(wh_hidden_dentry); + goto out_free; + } + + if (wh_hidden_dentry->d_inode) { + /* We found a whiteout so lets give up. */ + dprint(PRINT_DEBUG, "whiteout found in %d\n", bindex); + if (S_ISREG(wh_hidden_dentry->d_inode->i_mode)) { + set_dbend(dentry, bindex); + set_dbopaque(dentry, bindex); + DPUT(wh_hidden_dentry); + break; + } + err = -EIO; + printk(KERN_NOTICE "EIO: Invalid whiteout entry type" + " %d.\n", wh_hidden_dentry->d_inode->i_mode); + DPUT(wh_hidden_dentry); + DPUT(first_hidden_dentry); + goto out_free; + } + + DPUT(wh_hidden_dentry); + wh_hidden_dentry = NULL; + + /* Now do regular lookup; lookup foo */ + nd->path.dentry = dtohd_index(dentry, bindex); + /* FIXME: fix following line for mount point crossing */ + nd->path.mnt = stohiddenmnt_index(parent_dentry->d_sb, bindex); + + hidden_dentry = LOOKUP_ONE_LEN(name, hidden_dir_dentry, namelen); + print_dentry("hidden result", hidden_dentry); + if (IS_ERR(hidden_dentry)) { + DPUT(first_hidden_dentry); + err = PTR_ERR(hidden_dentry); + goto out_free; + } + + /* Store the first negative dentry specially, because if they + * are all negative we need this for future creates. */ + if (!hidden_dentry->d_inode) { + if (!first_hidden_dentry && (dbstart(dentry) == -1)) { + first_hidden_dentry = hidden_dentry; + first_dentry_offset = bindex; + } else { + DPUT(hidden_dentry); + } + continue; + } + + /* number of positive dentries */ + dentry_count++; + + /* store underlying dentry */ + if (dbstart(dentry) == -1) + set_dbstart(dentry, bindex); + set_dtohd_index(dentry, bindex, hidden_dentry); + set_dbend(dentry, bindex); + + /* update parent directory's atime with the bindex */ + fist_copy_attr_atime(parent_dentry->d_inode, + hidden_dir_dentry->d_inode); + + /* We terminate file lookups here. */ + if (!S_ISDIR(hidden_dentry->d_inode->i_mode)) { + if (lookupmode == INTERPOSE_PARTIAL) + continue; + if (dentry_count == 1) + goto out_positive; + /* This can only happen with mixed D-*-F-* */ + BUG_ON(!S_ISDIR(dtohd(dentry)->d_inode->i_mode)); + continue; + } + + opaque = is_opaque_dir(dentry, bindex); + if (opaque < 0) { + DPUT(first_hidden_dentry); + err = opaque; + goto out_free; + } + if (opaque) { + set_dbend(dentry, bindex); + set_dbopaque(dentry, bindex); + break; + } + } + + if (dentry_count) + goto out_positive; + else + goto out_negative; + + out_negative: + if (lookupmode == INTERPOSE_PARTIAL) + goto out; + + /* If we've only got negative dentries, then use the leftmost one. */ + if (lookupmode == INTERPOSE_REVAL) { + if (dentry->d_inode) { + itopd(dentry->d_inode)->uii_stale = 1; + } + goto out; + } + /* This should only happen if we found a whiteout. */ + if (first_dentry_offset == -1) { + nd->path.dentry = dentry; + /* FIXME: fix following line for mount point crossing */ + nd->path.mnt = stohiddenmnt_index(parent_dentry->d_sb, bindex); + + first_hidden_dentry = LOOKUP_ONE_LEN(name, hidden_dir_dentry, + namelen); + + first_dentry_offset = bindex; + if (IS_ERR(first_hidden_dentry)) { + err = PTR_ERR(first_hidden_dentry); + goto out; + } + } + set_dtohd_index(dentry, first_dentry_offset, first_hidden_dentry); + set_dbstart(dentry, first_dentry_offset); + set_dbend(dentry, first_dentry_offset); + + if (lookupmode == INTERPOSE_REVAL_NEG) + BUG_ON(dentry->d_inode != NULL); + else + d_add(dentry, NULL); + goto out; + +/* This part of the code is for positive dentries. */ + out_positive: + BUG_ON(dentry_count <= 0); + + /* If we're holding onto the first negative dentry throw it out. */ + DPUT(first_hidden_dentry); + + /* Partial lookups need to reinterpose, or throw away older negs. */ + if (lookupmode == INTERPOSE_PARTIAL) { + if (dentry->d_inode) { + unionfs_reinterpose(dentry); + goto out; + } + + /* This somehow turned positive, so it is as if we had a + * negative revalidation. */ + lookupmode = INTERPOSE_REVAL_NEG; + + update_bstart(dentry); + bstart = dbstart(dentry); + bend = dbend(dentry); + } + + err = unionfs_interpose(dentry, dentry->d_sb, lookupmode); + if (err) + goto out_drop; + + checkinode(dentry->d_inode, "unionfs_lookup OUT: child"); + checkinode(parent_dentry->d_inode, "unionfs_lookup OUT: dir"); + goto out; + + out_drop: + d_drop(dentry); + + out_free: + /* should dput all the underlying dentries on error condition */ + bstart = dbstart(dentry); + if (bstart >= 0) { + bend = dbend(dentry); + for (bindex = bstart; bindex <= bend; bindex++) + DPUT(dtohd_index(dentry, bindex)); + } + KFREE(dtohd_ptr(dentry)); + dtohd_ptr(dentry) = NULL; + set_dbstart(dentry, -1); + set_dbend(dentry, -1); + + out: + if (!err && dtopd(dentry)) { + BUG_ON(dbend(dentry) > dtopd(dentry)->udi_bcount); + BUG_ON(dbend(dentry) > sbmax(dentry->d_sb)); + BUG_ON(dbstart(dentry) < 0); + } + KFREE(whname); + print_dentry("OUT unionfs_lookup (parent)", parent_dentry); + print_dentry("OUT unionfs_lookup (child)", dentry); + if (locked_parent) + unlock_dentry(parent_dentry); + DPUT(parent_dentry); + if (locked_child) + unlock_dentry(dentry); + print_exit_status(err); + return ERR_PTR(err); +} + +/* This is a utility function that fills in a unionfs dentry.*/ +int unionfs_partial_lookup(struct dentry *dentry) +{ + struct dentry *tmp; + struct nameidata nd = { .flags = 0 }; + + tmp = unionfs_lookup_backend(dentry, &nd, INTERPOSE_PARTIAL); + if (!tmp) + return 0; + if (IS_ERR(tmp)) + return PTR_ERR(tmp); + /* need to change the interface */ + BUG_ON(tmp != dentry); + return -ENOSYS; +} + +/* The rest of these are utility functions for lookup. */ +static int is_opaque_dir(struct dentry *dentry, int bindex) +{ + int err = 0; + struct dentry *hidden_dentry; + struct dentry *wh_hidden_dentry; + struct inode *hidden_inode; + struct sioq_args args; + + print_entry_location(); + + hidden_dentry = dtohd_index(dentry, bindex); + hidden_inode = hidden_dentry->d_inode; + + BUG_ON(!S_ISDIR(hidden_inode->i_mode)); + + mutex_lock(&hidden_inode->i_mutex); + if (!inode_permission(hidden_inode, MAY_EXEC)) + wh_hidden_dentry = LOOKUP_ONE_LEN(UNIONFS_DIR_OPAQUE, + hidden_dentry, + sizeof(UNIONFS_DIR_OPAQUE) - 1); + else { + args.isopaque.dentry = hidden_dentry; + run_sioq(__is_opaque_dir, &args); + wh_hidden_dentry = args.ret; + } + + mutex_unlock(&hidden_inode->i_mutex); + if (IS_ERR(wh_hidden_dentry)) { + err = PTR_ERR(wh_hidden_dentry); + dprint(PRINT_DEBUG, "LOOKUP_ONE_LEN returned: %d\n", err); + goto out; + } + if (wh_hidden_dentry->d_inode) + err = 1; + DPUT(wh_hidden_dentry); + out: + print_exit_status(err); + return err; +} + +static int is_validname(const char *name) +{ + if (!strncmp(name, WHPFX, WHLEN)) + return 0; + if (!strncmp(name, UNIONFS_DIR_OPAQUE_NAME, + sizeof(UNIONFS_DIR_OPAQUE_NAME) - 1)) + return 0; + return 1; +} + +/* The dentry cache is just so we have properly sized dentries. */ +static struct kmem_cache *unionfs_dentry_cachep; +int init_dentry_cache(void) +{ + unionfs_dentry_cachep = + kmem_cache_create("unionfs_dentry", + sizeof(struct unionfs_dentry_info), 0, + SLAB_RECLAIM_ACCOUNT, NULL); + + if (!unionfs_dentry_cachep) + return -ENOMEM; + return 0; +} + +void destroy_dentry_cache(void) +{ + if (!unionfs_dentry_cachep) + return; + kmem_cache_destroy(unionfs_dentry_cachep); + return; +} + +void free_dentry_private_data(struct unionfs_dentry_info *udi) +{ + if (!udi) + return; + kmem_cache_free(unionfs_dentry_cachep, udi); +} + +int new_dentry_private_data(struct dentry *dentry) +{ + int newsize; + int oldsize = 0; + + spin_lock(&dentry->d_lock); + if (!dtopd_nocheck(dentry)) { + dtopd_lhs(dentry) = (struct unionfs_dentry_info *) + kmem_cache_alloc(unionfs_dentry_cachep, GFP_ATOMIC); + if (!dtopd_nocheck(dentry)) + goto out; +#ifdef CONFIG_PREEMPT_RT + init_MUTEX(&dtopd_nocheck(dentry)->udi_sem); + down(&dtopd_nocheck(dentry)->udi_sem); +#else + init_MUTEX_LOCKED(&dtopd_nocheck(dentry)->udi_sem); +#endif + +#ifdef TRACKLOCK + printk("INITLOCK:%p\n", dentry); +#endif + dtohd_ptr(dentry) = NULL; + } else { + oldsize = sizeof(struct dentry *) * dtopd(dentry)->udi_bcount; + } + + dtopd_nocheck(dentry)->udi_bstart = -1; + dtopd_nocheck(dentry)->udi_bend = -1; + dtopd_nocheck(dentry)->udi_bopaque = -1; + dtopd_nocheck(dentry)->udi_bcount = sbmax(dentry->d_sb); + atomic_set(&dtopd_nocheck(dentry)->udi_generation, + atomic_read(&stopd(dentry->d_sb)->usi_generation)); + newsize = sizeof(struct dentry *) * sbmax(dentry->d_sb); + + /* Don't reallocate when we already have enough space. */ + /* It would be ideal if we could actually use the slab macros to + * determine what our object sizes is, but those are not exported. + */ + if (oldsize) { + int minsize = 128; + + if (!newsize || ((oldsize < newsize) && (newsize > minsize))) { + KFREE(dtohd_ptr(dentry)); + dtohd_ptr(dentry) = NULL; + } + } + + if (!dtohd_ptr(dentry) && newsize) { + dtohd_ptr(dentry) = KMALLOC(newsize, GFP_ATOMIC); + if (!dtohd_ptr(dentry)) + goto out; + } + + if (oldsize > newsize) + memset(dtohd_ptr(dentry), 0, oldsize); + else + memset(dtohd_ptr(dentry), 0, newsize); + + spin_unlock(&dentry->d_lock); + return 0; + + out: + free_dentry_private_data(dtopd_nocheck(dentry)); + dtopd_lhs(dentry) = NULL; + spin_unlock(&dentry->d_lock); + return -ENOMEM; +} + +void update_bstart(struct dentry *dentry) +{ + int bindex; + int bstart = dbstart(dentry); + int bend = dbend(dentry); + struct dentry *hidden_dentry; + + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + if (hidden_dentry->d_inode) { + set_dbstart(dentry, bindex); + break; + } + DPUT(hidden_dentry); + set_dtohd_index(dentry, bindex, NULL); + } +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/main.c @@ -0,0 +1,869 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: main.c,v 1.176 2006/10/10 07:28:13 jsipek Exp $ + */ + +#include "unionfs.h" +#include <linux/module.h> +#include <linux/moduleparam.h> + +/* declarations added for "sparse" */ +extern void unionfs_kill_block_super(struct super_block *sb); + +/* declarations added for malloc_debugging */ + +#ifdef FIST_MALLOC_DEBUG +extern atomic_t unionfs_malloc_counter; +extern atomic_t unionfs_mallocs_outstanding; +#endif + +extern void unionfs_read_inode(struct inode *inode); + +static struct inode *unionfs_iget(struct super_block *sb, ino_t ino) +{ + struct inode *inode; + + inode = iget_locked(sb, ino); + if (!inode) + return ERR_PTR(-ENOMEM); + if (inode->i_state & I_NEW) { + unionfs_read_inode(inode); + unlock_new_inode(inode); + } + return inode; +} + +/* sb we pass is unionfs's super_block */ +int unionfs_interpose(struct dentry *dentry, struct super_block *sb, int flag) +{ + struct inode *hidden_inode; + struct dentry *hidden_dentry; + int err = 0; + struct inode *inode; + int is_negative_dentry = 1; + int bindex, bstart, bend; + + print_entry("flag = %d", flag); + + verify_locked(dentry); + + print_dentry("In unionfs_interpose", dentry); + + bstart = dbstart(dentry); + bend = dbend(dentry); + + /* Make sure that we didn't get a negative dentry. */ + for (bindex = bstart; bindex <= bend; bindex++) { + if (dtohd_index(dentry, bindex) && + dtohd_index(dentry, bindex)->d_inode) { + is_negative_dentry = 0; + break; + } + } + BUG_ON(is_negative_dentry); + + /* We allocate our new inode below, by calling iget. + * iget will call our read_inode which will initialize some + * of the new inode's fields + */ + + /* On revalidate we've already got our own inode and just need + * to fix it up. */ + if (flag == INTERPOSE_REVAL) { + inode = dentry->d_inode; + itopd(inode)->b_start = -1; + itopd(inode)->b_end = -1; + atomic_set(&itopd(inode)->uii_generation, + atomic_read(&stopd(sb)->usi_generation)); + + itohi_ptr(inode) = + KZALLOC(sbmax(sb) * sizeof(struct inode *), GFP_KERNEL); + if (!itohi_ptr(inode)) { + err = -ENOMEM; + goto out; + } + mutex_lock(&inode->i_mutex); + } else { + ino_t ino; + /* get unique inode number for unionfs */ +#ifdef UNIONFS_IMAP + if (stopd(sb)->usi_persistent) { + err = read_uin(sb, bindex, + dtohd_index(dentry, + bindex)->d_inode->i_ino, + O_CREAT, &ino); + if (err) + goto out; + } else +#endif + ino = iunique(sb, UNIONFS_ROOT_INO); + + inode = unionfs_iget(sb, ino); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + goto out; + } + + mutex_lock(&inode->i_mutex); + if (atomic_read(&inode->i_count) > 1) + goto skip; + } + + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) { + set_itohi_index(inode, bindex, NULL); + continue; + } + /* Initialize the hidden inode to the new hidden inode. */ + if (!hidden_dentry->d_inode) + continue; + set_itohi_index(inode, bindex, IGRAB(hidden_dentry->d_inode)); + } + + ibstart(inode) = dbstart(dentry); + ibend(inode) = dbend(dentry); + + /* Use attributes from the first branch. */ + hidden_inode = itohi(inode); + + /* Use different set of inode ops for symlinks & directories */ + if (S_ISLNK(hidden_inode->i_mode)) + inode->i_op = &unionfs_symlink_iops; + else if (S_ISDIR(hidden_inode->i_mode)) + inode->i_op = &unionfs_dir_iops; + + /* Use different set of file ops for directories */ + if (S_ISDIR(hidden_inode->i_mode)) + inode->i_fop = &unionfs_dir_fops; + + /* properly initialize special inodes */ + if (S_ISBLK(hidden_inode->i_mode) || S_ISCHR(hidden_inode->i_mode) || + S_ISFIFO(hidden_inode->i_mode) || S_ISSOCK(hidden_inode->i_mode)) + init_special_inode(inode, hidden_inode->i_mode, + hidden_inode->i_rdev); +#ifndef UNIONFS_MMAP + /* Fix our inode's address operations to that of the lower inode (Unionfs is FiST-Lite) */ + if (inode->i_mapping->a_ops != hidden_inode->i_mapping->a_ops) { + dprint(PRINT_DEBUG, "fixing inode 0x%p a_ops (0x%p -> 0x%p)\n", + inode, inode->i_mapping->a_ops, + hidden_inode->i_mapping->a_ops); + inode->i_mapping->a_ops = hidden_inode->i_mapping->a_ops; + } +#endif + /* all well, copy inode attributes */ + fist_copy_attr_all(inode, hidden_inode); + + skip: + /* only (our) lookup wants to do a d_add */ + switch (flag) { + case INTERPOSE_DEFAULT: + case INTERPOSE_REVAL_NEG: + d_instantiate(dentry, inode); + break; + case INTERPOSE_LOOKUP: + err = PTR_ERR(d_splice_alias(inode, dentry)); + break; + case INTERPOSE_REVAL: + /* Do nothing. */ + break; + default: + printk(KERN_ERR "Invalid interpose flag passed!"); + BUG(); + } + + print_dentry("Leaving unionfs_interpose", dentry); + print_inode("Leaving unionfs_interpose", inode); + mutex_unlock(&inode->i_mutex); + + out: + print_exit_status(err); + return err; +} + +void unionfs_reinterpose(struct dentry *dentry) +{ + struct dentry *hidden_dentry; + struct inode *inode; + int bindex, bstart, bend; + + print_entry_location(); + verify_locked(dentry); + print_dentry("IN: unionfs_reinterpose: ", dentry); + + /* This is pre-allocated inode */ + inode = dentry->d_inode; + + bstart = dbstart(dentry); + bend = dbend(dentry); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + + if (!hidden_dentry->d_inode) + continue; + if (itohi_index(inode, bindex)) + continue; + set_itohi_index(inode, bindex, IGRAB(hidden_dentry->d_inode)); + } + ibstart(inode) = dbstart(dentry); + ibend(inode) = dbend(dentry); + + print_dentry("OUT: unionfs_reinterpose: ", dentry); + print_inode("OUT: unionfs_reinterpose: ", inode); + + print_exit_location(); +} + +int check_branch(struct nameidata *nd) +{ + if (!strcmp(nd->path.dentry->d_sb->s_type->name, "unionfs")) + return -EINVAL; + if (!nd->path.dentry->d_inode) + return -ENOENT; + if (!S_ISDIR(nd->path.dentry->d_inode->i_mode)) + return -ENOTDIR; + return 0; +} + +/* checks if two hidden_dentries have overlapping branches */ +int is_branch_overlap(struct dentry *dent1, struct dentry *dent2) +{ + struct dentry *dent = NULL; + + dent = dent1; + while ((dent != dent2) && (dent->d_parent != dent)) { + dent = dent->d_parent; + } + if (dent == dent2) { + return 1; + } + + dent = dent2; + while ((dent != dent1) && (dent->d_parent != dent)) { + dent = dent->d_parent; + } + if (dent == dent1) { + return 1; + } + + return 0; +} +static int parse_branch_mode(char *name) +{ + int perms; + int l = strlen(name); + if (!strcmp(name + l - 3, "=ro")) { + perms = MAY_READ; + name[l - 3] = '\0'; + } else if (!strcmp(name + l - 6, "=nfsro")) { + perms = MAY_READ | MAY_NFSRO; + name[l - 6] = '\0'; + } else if (!strcmp(name + l - 3, "=rw")) { + perms = MAY_READ | MAY_WRITE; + name[l - 3] = '\0'; + } else { + perms = MAY_READ | MAY_WRITE; + } + + return perms; +} +static int get_separator_count(char *options, char *separator) +{ + char *token, *locopts, *locsep = NULL; + int count = 0; + /* + * We copy options so we dont destroy our pointer for parsing + */ + if (separator == NULL) { + locsep = KMALLOC(2, GFP_KERNEL); + if (!locsep) { + count = -ENOMEM; + goto out; + } + strcpy(locsep, ":"); + } else { + locsep = separator; + } + locopts = KMALLOC(strlen(options) + 1, GFP_KERNEL); + if (!locopts) { + count = -ENOMEM; + goto out; + } + strcpy(locopts, options); + while ((token = strsep(&locopts, locsep)) != NULL) + count++; + out: + KFREE(locopts); + return count; +} +static int parse_dirs_option(struct super_block *sb, struct unionfs_dentry_info + *hidden_root_info, char *options, char *separator) +{ + struct nameidata nd; + char *name, *locsep = NULL; + int err = 0; + int branches = 1; + int bindex = 0; + int i = 0; + int j = 0; + + struct dentry *dent1 = NULL; + struct dentry *dent2 = NULL; + + if (options[0] == '\0') { + printk(KERN_WARNING "unionfs: no branches specified\n"); + err = -EINVAL; + goto out; + } + /* + * Check to see if separator is specified otherwise use ':' + */ + if (separator == NULL) { + locsep = KMALLOC(2, GFP_KERNEL); + if (!locsep) { + err = -ENOMEM; + goto out; + } + strcpy(locsep, ":"); + } else { + locsep = separator; + } + branches = get_separator_count(options, separator); + /* allocate space for underlying pointers to hidden dentry */ + if (!(stopd(sb)->usi_data = alloc_new_data(branches))) { + err = -ENOMEM; + goto out; + } + + if (!(hidden_root_info->udi_dentry = alloc_new_dentries(branches))) { + err = -ENOMEM; + goto out; + } + + /* now parsing the string b1:b2=rw:b3=ro:b4 */ + branches = 0; + while ((name = strsep(&options, locsep)) != NULL) { + int perms; + + if (!*name) + continue; + branches++; + + /* strip off =rw or =ro if it is specified. */ + perms = parse_branch_mode(name); + if (!bindex && !(perms & MAY_WRITE)) { + err = -EINVAL; + goto out; + } + + dprint(PRINT_DEBUG, "using directory: %s (%c%c%c)\n", + name, perms & MAY_READ ? 'r' : '-', + perms & MAY_WRITE ? 'w' : '-', + perms & MAY_NFSRO ? 'n' : '-'); + + err = path_lookup(name, LOOKUP_FOLLOW, &nd); + RECORD_PATH_LOOKUP(&nd); + if (err) { + printk(KERN_WARNING "unionfs: error accessing " + "hidden directory '%s' (error %d)\n", name, err); + goto out; + } + + if ((err = check_branch(&nd))) { + printk(KERN_WARNING "unionfs: hidden directory " + "'%s' is not a valid branch\n", name); + path_put(&nd.path); + RECORD_PATH_RELEASE(&nd); + goto out; + } + + hidden_root_info->udi_dentry[bindex] = nd.path.dentry; + + set_stohiddenmnt_index(sb, bindex, nd.path.mnt); + set_branchperms(sb, bindex, perms); + set_branch_count(sb, bindex, 0); + + if (hidden_root_info->udi_bstart < 0) + hidden_root_info->udi_bstart = bindex; + hidden_root_info->udi_bend = bindex; + bindex++; + } + + if (branches == 0) { + printk(KERN_WARNING "unionfs: no branches specified\n"); + err = -EINVAL; + goto out; + } + + BUG_ON(branches != (hidden_root_info->udi_bend + 1)); + + /* ensure that no overlaps exist in the branches */ + for (i = 0; i < branches; i++) { + for (j = i + 1; j < branches; j++) { + dent1 = hidden_root_info->udi_dentry[i]; + dent2 = hidden_root_info->udi_dentry[j]; + + if (is_branch_overlap(dent1, dent2)) { + goto out_overlap; + } + } + } + + out_overlap: + + if (i != branches) { + printk(KERN_WARNING "unionfs: branches %d and %d overlap\n", i, + j); + err = -EINVAL; + goto out; + } + + out: + if (err) { + for (i = 0; i < branches; i++) { + if (hidden_root_info->udi_dentry[i]) + DPUT(hidden_root_info->udi_dentry[i]); + } + + KFREE(hidden_root_info->udi_dentry); + KFREE(stopd(sb)->usi_data); + + /* MUST clear the pointers to prevent potential double free if + * the caller dies later on + */ + hidden_root_info->udi_dentry = NULL; + stopd(sb)->usi_data = NULL; + } + if (!separator) + KFREE(locsep); + return err; +} + +/* + * Parse mount options. See the manual page for usage instructions. + * + * Returns the dentry object of the lower-level (hidden) directory; + * We want to mount our stackable file system on top of that hidden directory. + * + * Sets default debugging level to N, if any. + */ +static struct unionfs_dentry_info *unionfs_parse_options(struct super_block *sb, + char *options) +{ + struct unionfs_dentry_info *hidden_root_info; + char *optname, *separator = NULL; + int err = 0; + int bindex; + int sepfound = 0; + int dirsfound = 0; +#ifdef UNIONFS_IMAP + int imapfound = 0; +#endif + print_entry_location(); + + /* allocate private data area */ + err = -ENOMEM; + hidden_root_info = + KZALLOC(sizeof(struct unionfs_dentry_info), GFP_KERNEL); + if (!hidden_root_info) + goto out_error; + hidden_root_info->udi_bstart = -1; + hidden_root_info->udi_bend = -1; + hidden_root_info->udi_bopaque = -1; + + while ((optname = strsep(&options, ",")) != NULL) { + char *optarg; + char *endptr; + int intval; + + if (!*optname) { + continue; + } + + optarg = strchr(optname, '='); + if (optarg) { + *optarg++ = '\0'; + } + + /* All of our options take an argument now. Insert ones that + * don't, above this check. */ + if (!optarg) { + printk("unionfs: %s requires an argument.\n", optname); + err = -EINVAL; + goto out_error; + } + + if (!strcmp("dirs", optname)) { + if (++dirsfound > 1) { + printk(KERN_WARNING + "unionfs: multiple dirs specified\n"); + err = -EINVAL; + goto out_error; + } + err = + parse_dirs_option(sb, hidden_root_info, optarg, + separator); + if (err) + goto out_error; + continue; + } +#ifdef UNIONFS_IMAP + if (!strcmp("imap", optname)) { + if (++imapfound > 1) { + printk(KERN_WARNING + "unionfs: multiple imap specified\n"); + err = -EINVAL; + goto out_error; + } + err = parse_imap_option(sb, hidden_root_info, optarg); + if (err) + goto out_error; + continue; + } +#endif + if (!strcmp("delete", optname)) { + if (!strcmp("whiteout", optarg)) { + /* default */ +#ifdef UNIONFS_DELETE_ALL + } else if (!strcmp("all", optarg)) { + MOUNT_FLAG(sb) |= DELETE_ALL; +#endif + } else { + printk(KERN_WARNING + "unionfs: invalid delete option '%s'\n", + optarg); + err = -EINVAL; + goto out_error; + } + continue; + } + + if (!strcmp("separator", optname)) { + if (dirsfound) { + printk(KERN_WARNING + "unionfs: dirs= already parsed separator '%s' will have no effect\n", + optarg); + continue; + } + sepfound = 1; + separator = KMALLOC(strlen(optarg) + 1, GFP_KERNEL); + if (!separator) { + err = -ENOMEM; + goto out_error; + } + strcpy(separator, optarg); + continue; + } + /* All of these options require an integer argument. */ + intval = simple_strtoul(optarg, &endptr, 0); + if (*endptr) { + printk(KERN_WARNING + "unionfs: invalid %s option '%s'\n", + optname, optarg); + err = -EINVAL; + goto out_error; + } + + if (!strcmp("debug", optname)) { + set_debug_mask(intval); + continue; + } + + err = -EINVAL; + printk(KERN_WARNING + "unionfs: unrecognized option '%s'\n", optname); + goto out_error; + } + if (dirsfound != 1) { + printk(KERN_WARNING "unionfs: dirs option required\n"); + err = -EINVAL; + goto out_error; + } + goto out; + + out_error: + if (hidden_root_info && hidden_root_info->udi_dentry) { + for (bindex = hidden_root_info->udi_bstart; + bindex >= 0 && bindex <= hidden_root_info->udi_bend; + bindex++) { + struct dentry *d; + d = hidden_root_info->udi_dentry[bindex]; + DPUT(d); + if (stohiddenmnt_index(sb, bindex)) + mntput(stohiddenmnt_index(sb, bindex)); + } + } + + KFREE(hidden_root_info->udi_dentry); + KFREE(hidden_root_info); + + KFREE(stopd(sb)->usi_data); + stopd(sb)->usi_data = NULL; + + hidden_root_info = ERR_PTR(err); + KFREE(separator); + out: + print_exit_location(); + return hidden_root_info; +} + +static struct dentry *unionfs_d_alloc_root(struct super_block *sb) +{ + struct dentry *ret = NULL; + + if (sb) { + static const struct qstr name = {.name = "/",.len = 1 }; + + ret = d_alloc(NULL, &name); + if (ret) { + ret->d_op = &unionfs_dops; + ret->d_sb = sb; + ret->d_parent = ret; + } + } + return ret; +} + +static int unionfs_read_super(struct super_block *sb, void *raw_data, + int silent) +{ + int err = 0; + + struct unionfs_dentry_info *hidden_root_info = NULL; + int bindex, bstart, bend; + unsigned long long maxbytes; + + print_entry_location(); + + if (!raw_data) { + printk(KERN_WARNING + "unionfs_read_super: missing data argument\n"); + err = -EINVAL; + goto out; + } + + /* + * Allocate superblock private data + */ + stopd_lhs(sb) = KZALLOC(sizeof(struct unionfs_sb_info), GFP_KERNEL); + if (!stopd(sb)) { + printk(KERN_WARNING "%s: out of memory\n", __FUNCTION__); + err = -ENOMEM; + goto out; + } + stopd(sb)->b_end = -1; + atomic_set(&stopd(sb)->usi_generation, 1); + init_rwsem(&stopd(sb)->usi_rwsem); + + hidden_root_info = unionfs_parse_options(sb, raw_data); + if (IS_ERR(hidden_root_info)) { + printk(KERN_WARNING + "unionfs_read_super: error while parsing options (err = %ld)\n", + PTR_ERR(hidden_root_info)); + err = PTR_ERR(hidden_root_info); + hidden_root_info = NULL; + goto out_free; + } + if (hidden_root_info->udi_bstart == -1) { + err = -ENOENT; + goto out_free; + } + + /* set the hidden superblock field of upper superblock */ + bstart = hidden_root_info->udi_bstart; + BUG_ON(bstart != 0); + sbend(sb) = bend = hidden_root_info->udi_bend; + for (bindex = bstart; bindex <= bend; bindex++) { + struct dentry *d; + + d = hidden_root_info->udi_dentry[bindex]; + + set_stohs_index(sb, bindex, d->d_sb); + } + + /* Unionfs: Max Bytes is the maximum bytes from among all the branches */ + maxbytes = -1; + for (bindex = bstart; bindex <= bend; bindex++) + if (maxbytes < stohs_index(sb, bindex)->s_maxbytes) + maxbytes = stohs_index(sb, bindex)->s_maxbytes; + sb->s_maxbytes = maxbytes; + + sb->s_op = &unionfs_sops; +#ifdef CONFIG_EXPORTFS + sb->s_export_op = &unionfs_export_ops; +#endif + + /* + * we can't use d_alloc_root if we want to use + * our own interpose function unchanged, + * so we simply call our own "fake" d_alloc_root + */ + sb->s_root = unionfs_d_alloc_root(sb); + if (!sb->s_root) { + err = -ENOMEM; + goto out_dput; + } + + /* link the upper and lower dentries */ + dtopd_lhs(sb->s_root) = NULL; + if ((err = new_dentry_private_data(sb->s_root))) + goto out_freedpd; + + /* Set the hidden dentries for s_root */ + for (bindex = bstart; bindex <= bend; bindex++) { + struct dentry *d; + + d = hidden_root_info->udi_dentry[bindex]; + + set_dtohd_index(sb->s_root, bindex, d); + } + set_dbstart(sb->s_root, bstart); + set_dbend(sb->s_root, bend); + + /* Set the generation number to one, since this is for the mount. */ + atomic_set(&dtopd(sb->s_root)->udi_generation, 1); + + /* call interpose to create the upper level inode */ + if ((err = unionfs_interpose(sb->s_root, sb, 0))) + goto out_freedpd; + unlock_dentry(sb->s_root); + goto out; + + out_freedpd: + if (dtopd(sb->s_root)) { + KFREE(dtohd_ptr(sb->s_root)); + free_dentry_private_data(dtopd(sb->s_root)); + } + DPUT(sb->s_root); + out_dput: + if (hidden_root_info && !IS_ERR(hidden_root_info)) { + for (bindex = hidden_root_info->udi_bstart; + bindex <= hidden_root_info->udi_bend; bindex++) { + struct dentry *d; + + d = hidden_root_info->udi_dentry[bindex]; + + if (d) + DPUT(d); + + if (stopd(sb) && stohiddenmnt_index(sb, bindex)) + mntput(stohiddenmnt_index(sb, bindex)); + } + KFREE(hidden_root_info->udi_dentry); + KFREE(hidden_root_info); + hidden_root_info = NULL; + } + out_free: + KFREE(stopd(sb)->usi_data); + KFREE(stopd(sb)); + stopd_lhs(sb) = NULL; + out: + if (hidden_root_info && !IS_ERR(hidden_root_info)) { + KFREE(hidden_root_info->udi_dentry); + KFREE(hidden_root_info); + } + print_exit_status(err); + return err; +} + +static int unionfs_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, + void *raw_data, struct vfsmount *mnt) +{ + return get_sb_nodev(fs_type, flags, raw_data, unionfs_read_super, mnt); +} + +static struct file_system_type unionfs_fs_type = { + .owner = THIS_MODULE, + .name = "unionfs", + .get_sb = unionfs_get_sb, + .kill_sb = kill_anon_super, + .fs_flags = FS_REVAL_DOT, +}; + +static int init_debug = 0; +module_param_named(debug, init_debug, int, S_IRUGO); +MODULE_PARM_DESC(debug, "Initial Unionfs debug value."); + +static int __init init_unionfs_fs(void) +{ + int err; + printk("Registering unionfs " UNIONFS_VERSION "\n"); + + set_debug_mask(init_debug); + +#ifdef FIST_MALLOC_DEBUG + atomic_set(&unionfs_malloc_counter, 0); + atomic_set(&unionfs_mallocs_outstanding, 0); +#endif /* FIST_MALLOC_DEBUG */ + + if ((err = init_filldir_cache())) + goto out; + if ((err = init_inode_cache())) + goto out; + if ((err = init_dentry_cache())) + goto out; + if ((err = init_sioq())) + goto out; + err = register_filesystem(&unionfs_fs_type); + out: + if (err) { + fin_sioq(); + destroy_filldir_cache(); + destroy_inode_cache(); + destroy_dentry_cache(); + } + return err; +} +static void __exit exit_unionfs_fs(void) +{ + fin_sioq(); + destroy_filldir_cache(); + destroy_inode_cache(); + destroy_dentry_cache(); + unregister_filesystem(&unionfs_fs_type); + printk("Completed unionfs module unload.\n"); +} + +MODULE_AUTHOR + ("Filesystems and Storage Lab, Stony Brook University (http://www.fsl.cs.sunysb.edu/)"); +MODULE_DESCRIPTION("Unionfs " UNIONFS_VERSION + " (http://unionfs.filesystems.org/)"); +MODULE_LICENSE("GPL"); + +module_init(init_unionfs_fs); +module_exit(exit_unionfs_fs); +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/persistent_inode.c @@ -0,0 +1,658 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef Sipek + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: persistent_inode.c,v 1.36 2006/07/08 17:58:31 ezk Exp $ + */ +#ifdef UNIONFS_IMAP + +#include "unionfs.h" + +static ssize_t __fread(struct file *filp, void *buf, size_t size, loff_t * pos) +{ + int err; + mm_segment_t oldfs; + ssize_t(*func) (struct file *, char __user *, size_t, loff_t *); + + func = do_sync_read; + if (filp->f_op && filp->f_op->read) + func = filp->f_op->read; + + oldfs = get_fs(); + set_fs(KERNEL_DS); + do { + err = func(filp, (char __user *)buf, size, pos); + } while (err == -EAGAIN || err == -EINTR); + set_fs(oldfs); + return err; +} + +static ssize_t __fwrite(struct file *filp, void *buf, size_t size, loff_t * pos) +{ + int err; + mm_segment_t oldfs; + unsigned long flim; + struct rlimit *rl; + ssize_t(*func) (struct file *, const char __user *, size_t, loff_t *); + + func = do_sync_write; + if (filp->f_op && filp->f_op->write) + func = filp->f_op->write; + + /* + * it breaks RLIMIT_FSIZE, + * but users should be careful to quota. + */ + rl = current->signal->rlim + RLIMIT_FSIZE; + flim = rl->rlim_cur; + rl->rlim_cur = RLIM_INFINITY; + oldfs = get_fs(); + set_fs(KERNEL_DS); + do { + err = func(filp, (const char __user *)buf, size, pos); + } while (err == -EAGAIN || err == -EINTR); + set_fs(oldfs); + rl->rlim_cur = flim; + return err; +} + +/* + * verify_forwardmap(super_block *sb) + * sb: pointer to a superblock containing the forwardmap. + * returns: 0 on success EINVAL or ENOMEM on failure; + */ +static int verify_forwardmap(struct super_block *sb) +{ + int err = 0, bytesread = 0, bindex = 0, mallocsize = 0; + loff_t readpos = 0; + struct file *forwardmap = NULL; + struct fmaphdr header; + struct unionfs_sb_info *spd = NULL; + print_entry_location(); + + spd = stopd(sb); + BUG_ON(!spd); + + forwardmap = spd->usi_forwardmap; + if (!forwardmap) { + err = -EINVAL; + goto out; + } + bytesread = __fread(forwardmap, &header, sizeof(struct fmaphdr), + &readpos); + if (bytesread < sizeof(struct fmaphdr)) { + err = -EINVAL; + goto out; + } + if (header.magic != FORWARDMAP_MAGIC + || header.version != FORWARDMAP_VERSION) { + err = -EINVAL; + goto out; + } + spd->usi_bmap = + KMALLOC(sizeof(struct bmapent) * header.usedbranches, GFP_KERNEL); + + if (!spd->usi_bmap) { + err = -ENOMEM; + goto out; + } + + while (bindex < header.usedbranches) { + bytesread = __fread(forwardmap, &stopd(sb)->usi_bmap[bindex], + sizeof(struct bmapent), &readpos); + if (bytesread < sizeof(struct bmapent)) { + err = -EINVAL; + goto out_err; + } + bindex++; + } + + mallocsize = sizeof(int) * header.usedbranches; + goto out; + out_err: + if (spd->usi_bmap) + KFREE(spd->usi_bmap); + out: + print_exit_status(err); + return err; +} + +/* + * verify_reversemap(struct super_block sb, int rmapindex) + * + * sb: The unionfs superblock containing all of the current imap info + * rmapindex: the index in the usi_reversemaps array that we wish to + * verify + * + * Assumes the reverse maps less than rmapindex are valid. + * + * returns: 0 if the opperation succeds + * -EINVAL if the map file does not belong to the forward map + * + */ +static int verify_reversemap(struct super_block *sb, int rmapindex, + struct unionfs_dentry_info *hidden_root_info) +{ + int err = 0, i = 0, bindex = 0, found = 0, bytesread; + loff_t readpos = 0; + struct file *forwardmap, *reversemap; + struct fmaphdr fheader; + struct rmaphdr rheader; + struct kstatfs st; + struct unionfs_sb_info *spd = NULL; + + print_entry_location(); + + spd = stopd(sb); + BUG_ON(!spd); + + forwardmap = spd->usi_forwardmap; + if (!forwardmap) { + err = -EINVAL; + goto out; + } + reversemap = spd->usi_reversemaps[rmapindex]; + if (!reversemap) { + err = -EINVAL; + goto out; + } + bytesread = __fread(forwardmap, &fheader, sizeof(struct fmaphdr), + &readpos); + if (bytesread < sizeof(struct fmaphdr)) { + err = -EINVAL; + goto out; + } + readpos = 0; + bytesread = __fread(reversemap, &rheader, sizeof(struct rmaphdr), + &readpos); + if (bytesread < sizeof(struct rmaphdr)) { + err = -EINVAL; + goto out; + } + if (rheader.magic != REVERSEMAP_MAGIC + || rheader.version != REVERSEMAP_VERSION) { + err = -EINVAL; + goto out; + } + if (memcmp(fheader.uuid, rheader.fwduuid, sizeof(fheader.uuid))) { + err = -EINVAL; + goto out; + } + + /* XXX: Ok so here we take the new map and read the fsid from it. Then + * we go through all the branches in the union and see which ones it + * matches with*/ + for (i = 0; i < spd->usi_num_bmapents && !found; i++) { + if (memcmp + (rheader.revuuid, spd->usi_bmap[i].uuid, + sizeof(rheader.revuuid))) + continue; + + found = 1; + for (bindex = 0; bindex <= hidden_root_info->udi_bend; bindex++) { + struct dentry *d; + fsid_t fsid; + dev_t dev; + memset(&st, 0, sizeof(struct kstatfs)); + + d = hidden_root_info->udi_dentry[bindex]; + + err = d->d_sb->s_op->statfs(d->d_sb, &st); + if (err) + goto out; + + if (st.f_fsid.val[0] || st.f_fsid.val[1]) { + fsid = st.f_fsid; + } else { + + dev = d->d_sb->s_dev; + fsid.val[0] = MAJOR(dev); + fsid.val[1] = MINOR(dev); + } + + if (memcmp(&fsid, &rheader.fsid, sizeof(fsid))) + continue; + + if (spd->usi_bnum_table[bindex] == -1) + spd->usi_bnum_table[bindex] = i; + if (spd->usi_map_table[bindex]) { + printk(KERN_WARNING + "Two reverse maps share fsid %u%u!\n", + rheader.fsid.val[0], + rheader.fsid.val[1]); + err = -EINVAL; + goto out; + } else { + spd->usi_map_table[bindex] = reversemap; + } + } + } + if (!found) { + printk(KERN_WARNING + "Could not match the reversemap uuid with an entry in the forwardmap table\n"); + err = -EINVAL; + } + out: + print_exit_status(err); + return err; +} + +int init_imap_data(struct super_block *sb, + struct unionfs_dentry_info *hidden_root_info) +{ + int i, err = 0, mallocsize = 0; + struct unionfs_sb_info *spd; + + print_entry_location(); + + spd = stopd(sb); + + spd->usi_forwardmap = NULL; + spd->usi_reversemaps = NULL; + spd->usi_bnum_table = NULL; + + mallocsize = sizeof(struct file *) * (hidden_root_info->udi_bend + 1); + spd->usi_reversemaps = KZALLOC(mallocsize, GFP_KERNEL); + if (!spd->usi_reversemaps) { + err = -ENOMEM; + goto out_error; + } + + spd->usi_map_table = KZALLOC(mallocsize, GFP_KERNEL); + if (!spd->usi_map_table) { + err = -ENOMEM; + goto out_error; + } + + mallocsize = sizeof(int) * (hidden_root_info->udi_bend + 1); + spd->usi_bnum_table = KMALLOC(mallocsize, GFP_KERNEL); + if (!spd->usi_bnum_table) { + err = -ENOMEM; + goto out_error; + } + + for (i = 0; i <= hidden_root_info->udi_bend; i++) { + spd->usi_bnum_table[i] = -1; + } + + if (!err) + goto out; + out_error: + + if (spd->usi_reversemaps) { + KFREE(spd->usi_reversemaps); + spd->usi_reversemaps = NULL; + } + + if (spd->usi_map_table) { + KFREE(spd->usi_map_table); + spd->usi_map_table = NULL; + } + + if (spd->usi_bnum_table) { + KFREE(spd->usi_bnum_table); + spd->usi_bnum_table = NULL; + + } + + out: + print_exit_status(err); + return err; + +} + +void cleanup_imap_data(struct super_block *sb) +{ + int count = 0; + struct unionfs_sb_info *spd; + + print_entry_location(); + + spd = stopd(sb); + + spd->usi_persistent = 0; + count = spd->usi_num_bmapents; + while (count - 1 >= 0) { + if (spd->usi_reversemaps[count - 1]) { + filp_close(spd->usi_reversemaps[count - 1], NULL); + spd->usi_reversemaps[count - 1] = NULL; + } + count--; + } + if (spd->usi_reversemaps) { + KFREE(spd->usi_reversemaps); + spd->usi_reversemaps = NULL; + } + + if (spd->usi_map_table) { + KFREE(spd->usi_map_table); + spd->usi_map_table = NULL; + } + + if (spd->usi_bnum_table) { + KFREE(spd->usi_bnum_table); + spd->usi_bnum_table = NULL; + } + if (spd->usi_forwardmap) { + filp_close(spd->usi_forwardmap, NULL); + spd->usi_forwardmap = NULL; + } + print_exit_location(); +} + +int parse_imap_option(struct super_block *sb, + struct unionfs_dentry_info *hidden_root_info, + char *options) +{ + int count = 0, err = 0; + char *name; + struct unionfs_sb_info *spd = NULL; + + print_entry_location(); + spd = stopd(sb); + BUG_ON(!spd); + + err = init_imap_data(sb, hidden_root_info); + if (err) + goto out_error; + while ((name = strsep(&options, ":")) != NULL) { + if (!*name) + continue; + if (!spd->usi_forwardmap) { + spd->usi_forwardmap = filp_open(name, O_RDWR, 0); + if (IS_ERR(spd->usi_forwardmap)) { + err = PTR_ERR(spd->usi_forwardmap); + spd->usi_forwardmap = NULL; + goto out_error; + } + } else { + spd->usi_reversemaps[count] = + filp_open(name, O_RDWR, 0); + if (IS_ERR(spd->usi_reversemaps[count])) { + err = PTR_ERR(spd->usi_reversemaps[count]); + spd->usi_reversemaps[count] = NULL; + goto out_error; + + } + count++; + } + } + if (count <= 0) { + printk(KERN_WARNING "unionfs: no reverse maps specified.\n"); + err = -EINVAL; + } + if (err) + goto out_error; + + /* Initialize the super block's next_avail field */ + /* Dave, you can't use 64-bit division here because the i386 doesn't + * support it natively. Instead you need to punt if the size is + * greater than unsigned long, and then cast it down. Then you should + * be able to assign to this value, without having these problems. */ + + if (spd->usi_forwardmap->f_dentry->d_inode->i_size > ULONG_MAX) { + err = -EFBIG; + goto out_error; + } + spd->usi_next_avail = + ((unsigned long)(spd->usi_forwardmap->f_dentry->d_inode-> + i_size - (sizeof(struct fmaphdr) + + sizeof(struct bmapent[256]))) + / sizeof(struct fmapent)); + + if (spd->usi_next_avail < FIRST_VALID_INODE) + spd->usi_next_avail = FIRST_VALID_INODE; + + spd->usi_num_bmapents = count; + err = verify_forwardmap(sb); + if (err) + goto out_error; + while (count > 0) { + err = verify_reversemap(sb, --count, hidden_root_info); + if (err) + goto out_error; + } + spd->usi_persistent = 1; + + goto out; + + out_error: + spd->usi_num_bmapents = count; + cleanup_imap_data(sb); + + out: + print_exit_status(err); + return err; +} + + /* + * get @ino from @hidden_ino. + */ +static int __read_uin(struct unionfs_sb_info *sbi, ino_t hidden_ino, int bindex, + ino_t * ino) +{ + int err; + struct file *rev; + loff_t pos; + ssize_t sz; + uint64_t ino64; + const int elmnt = sizeof(ino64); + + rev = sbi->usi_map_table[bindex]; + pos = sizeof(struct rmaphdr) + elmnt * hidden_ino; + *ino = 0; + err = 0; + if (pos + elmnt > rev->f_dentry->d_inode->i_size) + goto out; + + sz = __fread(rev, &ino64, elmnt, &pos); + err = sz; + if (err < 0) + goto out; + err = 0; + *ino = -1; + if (sz != elmnt || ino64 > *ino) + err = -EIO; + *ino = ino64; + out: + print_exit_status(err); + return err; +} + +/* + * put unionfs @ino for @hidden_ino on @bindex. + */ +static int __write_uin(struct unionfs_sb_info *sbi, ino_t ino, int bindex, + ino_t hidden_ino) +{ + struct file *fwd, *rev; + struct fmapent ent; + loff_t pos; + ssize_t sz; + int err; + uint64_t ino64; + const int fwdhdr = sizeof(struct fmaphdr) + sizeof(struct bmapent[256]); + const int fwd_elmnt = sizeof(ent); + const int rev_elmnt = sizeof(ino64); + + err = -ENOSPC; + if (ino < FIRST_VALID_INODE) + goto out; + + fwd = sbi->usi_forwardmap; + ent.fsnum = sbi->usi_bnum_table[bindex]; + ent.inode = hidden_ino; + pos = fwdhdr + fwd_elmnt * ino; + sz = __fwrite(fwd, &ent, fwd_elmnt, &pos); + err = sz; + if (err < 0) + goto out; + err = -EIO; + if (sz != fwd_elmnt) + goto out; + + rev = sbi->usi_map_table[bindex]; + pos = sizeof(struct rmaphdr) + rev_elmnt * hidden_ino; + ino64 = ino; + sz = __fwrite(rev, &ino64, rev_elmnt, &pos); + err = sz; + if (err < 0) + goto out; + err = 0; + if (sz != rev_elmnt) + err = -EIO; + out: + print_exit_status(err); + return err; +} + +/* + * read_uin(struct super_block *sb, uint8_t branchnum, ino_t inode_number, int flag, ino_t *uino) + * fsnum: branch to reference when getting the inode number + * inode_number: lower level inode number use to reference the proper inode. + * flag: if set to O_CREAT it will creat the entry if it doesent exist + * otherwise it will return the existing one. + * returns: the unionfs inode number either created or retrieved based on + * the information. + */ +int read_uin(struct super_block *sb, uint8_t branchnum, ino_t inode_number, + int flag, ino_t * uino) +{ + int err = 0; + struct unionfs_sb_info *spd; + + print_entry_location(); + + spd = stopd(sb); + BUG_ON(!spd); + + /* Find appropriate reverse map and then read from the required position */ + /* get it from the array. */ + err = __read_uin(spd, inode_number, branchnum, uino); + if (err || *uino) + goto out; + + err = -EIO; + if (!(flag & O_CREAT)) + goto out; + + /* If we haven't found an entry and we have the O_CREAT flag set we want to + * create a new entry write it out to the file and return its index + */ + mutex_lock(&sb->s_lock); + *uino = spd->usi_next_avail++; + err = __write_uin(spd, *uino, branchnum, inode_number); + if (err) + spd->usi_next_avail--; + mutex_unlock(&sb->s_lock); + out: + print_exit_status(err); + return err; +} + +int write_uin(struct super_block *sb, ino_t ino, int bindex, ino_t hidden_ino) +{ + int err; + + print_entry_location(); + err = __write_uin(stopd(sb), ino, bindex, hidden_ino); + print_exit_status(err); + return err; +} + +/* + * get_lin(ino_t inode_number) + * inode_number : inode number for the unionfs inode + * returns: the lower level inode# and branch# + */ +/* entry should use a poiner on the stack. should be staticly allocated one + * level up*/ +int get_lin(struct super_block *sb, ino_t inode_number, struct fmapent *entry) +{ + struct file *forwardmap; + loff_t seek_size; + mm_segment_t oldfs; + int err = 0, bytesread = 0; + + print_entry_location(); + + if (!entry) { + entry = ERR_PTR(-ENOMEM); + goto out; + } + forwardmap = stopd(sb)->usi_forwardmap; + seek_size = + sizeof(struct fmaphdr) + sizeof(struct bmapent[256]) + + (sizeof(struct fmapent) * inode_number); + oldfs = get_fs(); + set_fs(KERNEL_DS); + bytesread = __fread(forwardmap, entry, sizeof(*entry), &seek_size); + set_fs(oldfs); + if (bytesread != sizeof(*entry)) + err = -EINVAL; + + out: + print_exit_location(); + return err; +} + +/* + * remove_map(struct super_block *sb,int bindex) + * + * sb: The super block containing all the current imap info + * bindex: the index of the branch that is being removed. + * + * This assumes that end hasen't been decremented yet. + * + * Returns: This function really can't fail. The only thing + * that could possibly happen is that it will oops but that + * requires unionfs to be in an inconsistant state which + * shoulden't happen. + */ +int remove_map(struct super_block *sb, int bindex) +{ + int i; + struct unionfs_sb_info *spd; + + print_entry_location(); + + spd = stopd(sb); + BUG_ON(!spd); + + for (i = bindex; i < sbend(sb); i++) { + spd->usi_map_table[i] = spd->usi_map_table[i + 1]; + spd->usi_bnum_table[i] = spd->usi_bnum_table[i + 1]; + } + return 0; +} + +#endif +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/print.c @@ -0,0 +1,439 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef Sipek + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: print.c,v 1.77 2006/07/08 17:58:31 ezk Exp $ + */ + +/* Print debugging functions */ + +#include "unionfs.h" + +static unsigned int debug_mask = DEFAULT_DEBUG_MASK; + +/* get value of debugging variable */ +unsigned int get_debug_mask(void) +{ + return debug_mask; +} + +/* set debug level variable and return the previous value */ +int set_debug_mask(int val) +{ +#ifdef UNIONFS_DEBUG + int prev = debug_mask; + + debug_mask = val; + + printk(KERN_INFO UNIONFS_NAME ": debug mask set to %u\n", debug_mask); + + return prev; +#else /* UNIONFS_DEBUG */ + printk(KERN_WARNING UNIONFS_NAME ": debugging is not enabled\n"); + return -ENOTSUPP; +#endif /* ! UNIONFS_DEBUG */ +} + +static inline int should_print(const unsigned int req) +{ + return (req & debug_mask); +} + +static void unionfs_print_generic_inode(const char *prefix, + const char *prefix2, const struct inode *inode) +{ + if (!inode) { + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: NULL INODE PASSED!\n", prefix, prefix2); + return; + } + + if (IS_ERR(inode)) { + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: ERROR INODE PASSED: %ld\n", prefix, prefix2, + PTR_ERR(inode)); + return; + } + + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_ino=%lu\n", + prefix, prefix2, inode->i_ino); + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_count=%u\n", + prefix, prefix2, atomic_read(&inode->i_count)); + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_nlink=%u\n", + prefix, prefix2, inode->i_nlink); + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_mode=%o\n", + prefix, prefix2, inode->i_mode); + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_size=%llu\n", + prefix, prefix2, inode->i_size); + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_op=%p\n", + prefix, prefix2, inode->i_op); + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s%s: i_sb=%p (%s)\n", + prefix, prefix2, inode->i_sb, (inode->i_sb ? sbt(inode->i_sb) : "NullTypeSB")); +} + +void unionfs_print_inode(const unsigned int req, const char *prefix, const struct inode *inode) +{ + int bindex; + + if (!should_print(req)) + return; + + if (!inode) { + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s: NULL INODE PASSED!\n", prefix); + return; + } + if (IS_ERR(inode)) { + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s: ERROR INODE PASSED: %ld\n", prefix, PTR_ERR(inode)); + return; + } + + unionfs_print_generic_inode(prefix, "", inode); + + if (strcmp("unionfs", sbt(inode->i_sb))) { + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s: Not a " UNIONFS_NAME " inode.\n", prefix); + return; + } + + if (!itopd(inode)) + return; + + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s: ibstart=%d, ibend=%d\n", prefix, ibstart(inode), ibend(inode)); + + if (ibstart(inode) == -1) + return; + + for (bindex = ibstart(inode); bindex <= ibend(inode); bindex++) { + struct inode *hidden_inode = itohi_index(inode, bindex); + char newstr[10]; + if (!hidden_inode) { + printk(KERN_DEBUG UNIONFS_NAME ": PI:%s: HI#%d: NULL\n", prefix, bindex); + continue; + } + snprintf(newstr, 10, ": HI%d", bindex); + unionfs_print_generic_inode(prefix, newstr, hidden_inode); + } +} + +static void unionfs_print_generic_file(const char *prefix, const char *prefix2, + const struct file *file) +{ + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_dentry=0x%p\n", prefix, prefix2, file->f_dentry); + + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: name=%s\n", prefix, prefix2, file->f_dentry->d_name.name); + if (file->f_dentry->d_inode) { + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_dentry->d_inode->i_ino=%lu\n", prefix, prefix2, file->f_dentry->d_inode->i_ino); + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_dentry->d_inode->i_mode=%o\n", prefix, prefix2, file->f_dentry->d_inode->i_mode); + } + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_op=0x%p\n", prefix, prefix2, file->f_op); + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_mode=0x%x\n", prefix, prefix2, file->f_mode); + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_pos=0x%llu\n", prefix, prefix2, file->f_pos); + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_count=%u\n", prefix, prefix2, atomic_read(&file->f_count)); + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_flags=0x%x\n", prefix, prefix2, file->f_flags); + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s%s: f_version=%llu\n", prefix, prefix2, file->f_version); +} + +void unionfs_print_file(const unsigned int req, const char *prefix, const struct file *file) +{ + struct file *hidden_file; + + if (!should_print(req)) + return; + + if (!file) { + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s: NULL FILE PASSED!\n", prefix); + return; + } + + unionfs_print_generic_file(prefix, "", file); + + if (strcmp("unionfs", sbt(file->f_dentry->d_sb))) { + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s: Not a " UNIONFS_NAME " file.\n", prefix); + return; + } + + if (ftopd(file)) { + int bindex; + + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s: fbstart=%d, fbend=%d\n", prefix, fbstart(file), fbend(file)); + + for (bindex = fbstart(file); bindex <= fbend(file); bindex++) { + char newstr[10]; + hidden_file = ftohf_index(file, bindex); + if (!hidden_file) { + printk(KERN_DEBUG UNIONFS_NAME ": PF:%s: HF#%d is NULL\n", prefix, bindex); + continue; + } + snprintf(newstr, 10, ": HF%d", bindex); + unionfs_print_generic_file(prefix, newstr, hidden_file); + } + } +} + +static char mode_to_type(mode_t mode) +{ + if (S_ISDIR(mode)) + return 'd'; + if (S_ISLNK(mode)) + return 'l'; + if (S_ISCHR(mode)) + return 'c'; + if (S_ISBLK(mode)) + return 'b'; + if (S_ISREG(mode)) + return 'f'; + return '?'; +} + +static void unionfs_print_generic_dentry(const char *prefix, const char *prefix2, const + struct dentry *dentry, int check) +{ + if (!dentry) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: NULL DENTRY PASSED!\n", prefix, prefix2); + return; + } + + if (IS_ERR(dentry)) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: ERROR DENTRY (%ld)!\n", prefix, prefix2, + PTR_ERR(dentry)); + return; + } + + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: dentry = %p\n", prefix, prefix2, dentry); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_count=%d\n", prefix, prefix2, atomic_read(&dentry->d_count)); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_flags=%x\n", prefix, prefix2, (int)dentry->d_flags); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_name.name=\"%s\" (len = %d)\n", prefix, prefix2, dentry->d_name.name, dentry->d_name.len); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_sb=%p (%s)\n", prefix, prefix2, dentry->d_sb, sbt(dentry->d_sb)); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_inode=%p\n", prefix, prefix2, dentry->d_inode); + + if (dentry->d_inode) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_inode->i_ino=%ld (%s)\n", prefix, prefix2, + dentry->d_inode->i_ino, + sbt(dentry->d_inode->i_sb)); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: dentry->d_inode->i_mode: %c%o\n", prefix, + prefix2, mode_to_type(dentry->d_inode->i_mode), + dentry->d_inode->i_mode); + } + + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_parent=%p (%s)\n", prefix, prefix2, + dentry->d_parent, + (dentry->d_parent ? sbt(dentry->d_parent->d_sb) : "nil")); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_parent->d_name.name=\"%s\"\n", prefix, prefix2, + dentry->d_parent->d_name.name); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_parent->d_count=%d\n", prefix, prefix2, + atomic_read(&dentry->d_parent->d_count)); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_op=%p\n", prefix, prefix2, dentry->d_op); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: d_fsdata=%p\n", prefix, prefix2, + dentry->d_fsdata); + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s%s: hlist_unhashed(d_hash)=%d\n", prefix, prefix2, + hlist_unhashed(&((struct dentry *)dentry)->d_hash)); + + /* After we have printed it, we can assert something about it. */ + if (check) + BUG_ON(atomic_read(&dentry->d_count) <= 0); +} + +static void __unionfs_print_dentry(const char *prefix, const struct dentry *dentry, + int check) +{ + if (!dentry) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s: NULL DENTRY PASSED!\n", prefix); + return; + } + + if (IS_ERR(dentry)) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s: ERROR DENTRY (%ld)!\n", prefix, + PTR_ERR(dentry)); + return; + } + + unionfs_print_generic_dentry(prefix, "", dentry, check); + + if (strcmp("unionfs", sbt(dentry->d_sb))) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s: Not a " UNIONFS_NAME " dentry.\n", prefix); + return; + } + + if (!dtopd(dentry)) + return; + + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s: dbstart=%d, dbend=%d, dbopaque=%d\n", + prefix, dbstart(dentry), dbend(dentry), dbopaque(dentry)); + + if (dbstart(dentry) != -1) { + int bindex; + char newstr[10]; + struct dentry *hidden_dentry; + + for (bindex = dbstart(dentry); bindex <= dbend(dentry); + bindex++) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) { + printk(KERN_DEBUG UNIONFS_NAME ": PD:%s: HD#%d: NULL\n", prefix, bindex); + continue; + } + snprintf(newstr, 10, ": HD%d", bindex); + unionfs_print_generic_dentry(prefix, newstr, hidden_dentry, check); + } + } +} + +void unionfs_print_dentry(const unsigned int req, const char *prefix, const struct dentry *dentry) +{ + if (!should_print(req)) + return; + + __unionfs_print_dentry(prefix, dentry, 1); +} + +void unionfs_print_dentry_nocheck(const unsigned int req, const char *prefix, const struct dentry *dentry) +{ + if (!should_print(req)) + return; + + __unionfs_print_dentry(prefix, dentry, 0); +} + +void unionfs_checkinode(const unsigned int req, const struct inode *inode, const char *msg) +{ + if (!should_print(req)) + return; + + if (!inode) { + printk(KERN_DEBUG UNIONFS_NAME ": unionfs_checkinode - inode is NULL! (%s)\n", + msg); + return; + } + + if (!itopd(inode)) { + printk(KERN_DEBUG UNIONFS_NAME ": unionfs_checkinode(%ld) - no private data (%s)\n", + inode->i_ino, msg); + return; + } + + if ((itopd(inode)->b_start < 0) || !itohi(inode)) { + printk(KERN_DEBUG UNIONFS_NAME + "unionfs_checkinode(%ld) - underlying is NULL! (%s)\n", + inode->i_ino, msg); + return; + } + + if (!inode->i_sb) { + printk(KERN_DEBUG UNIONFS_NAME + ": unionfs_checkinode(%ld) - inode->i_sb is NULL! (%s)\n", + inode->i_ino, msg); + return; + } + + printk(KERN_DEBUG UNIONFS_NAME ": inode->i_sb->s_type %p\n", inode->i_sb->s_type); + if (!inode->i_sb->s_type) { + printk(KERN_DEBUG UNIONFS_NAME + ": unionfs_checkinode(%ld) - inode->i_sb->s_type is NULL! (%s)\n", + inode->i_ino, msg); + return; + } + + printk(KERN_DEBUG UNIONFS_NAME + ": CI: %s: inode->i_count = %d, hidden_inode->i_count = %d, inode = %lu, sb = %s, hidden_sb = %s\n", + msg, atomic_read(&inode->i_count), + itopd(inode)->b_start >= + 0 ? atomic_read(&itohi(inode)->i_count) : -1, inode->i_ino, + inode->i_sb->s_type->name, + itopd(inode)->b_start >= + 0 ? itohi(inode)->i_sb->s_type->name : "(none)"); +} + +void unionfs_print_sb(const unsigned int req, const char *prefix, const struct super_block *sb) +{ + struct super_block *hidden_superblock; + + if (!should_print(req)) + return; + + if (!sb) { + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: NULL SB PASSED!\n", prefix); + return; + } + + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_blocksize=%lu\n", prefix, sb->s_blocksize); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_blocksize_bits=%u\n", prefix, sb->s_blocksize_bits); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_flags=0x%lx\n", prefix, sb->s_flags); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_magic=0x%lx\n", prefix, sb->s_magic); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_maxbytes=%llu\n", prefix, sb->s_maxbytes); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_count=%d\n", prefix, sb->s_count); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: s_active=%d\n", prefix, atomic_read(&sb->s_active)); + + if (stopd(sb)) + printk(KERN_DEBUG UNIONFS_NAME ": sbstart=%d, sbend=%d\n", sbstart(sb), + sbend(sb)); + + if (stopd(sb)) { + int bindex; + for (bindex = sbstart(sb); bindex <= sbend(sb); bindex++) { + hidden_superblock = stohs_index(sb, bindex); + if (!hidden_superblock) { + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d is NULL", prefix, + bindex); + continue; + } + + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_blocksize=%lu\n", prefix, bindex, + hidden_superblock->s_blocksize); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_blocksize_bits=%u\n", prefix, bindex, + hidden_superblock->s_blocksize_bits); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_flags=0x%lx\n", prefix, bindex, + hidden_superblock->s_flags); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_magic=0x%lx\n", prefix, bindex, + hidden_superblock->s_magic); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_maxbytes=%llu\n", prefix, bindex, + hidden_superblock->s_maxbytes); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_count=%d\n", prefix, bindex, + hidden_superblock->s_count); + printk(KERN_DEBUG UNIONFS_NAME ": PSB:%s: HS#%d: s_active=%d\n", prefix, bindex, + atomic_read(&hidden_superblock->s_active)); + } + } +} + +int unionfs_print(const unsigned int req, const char *fmt, ...) +{ + va_list ap; + int r; + + if (!should_print(req)) + return 0; + + printk(KERN_DEBUG UNIONFS_NAME ": "); + va_start(ap, fmt); + r = vprintk(fmt, ap); + va_end(ap); + + return r; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/rdstate.c @@ -0,0 +1,328 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: rdstate.c,v 1.34 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +/* This file contains the routines for maintaining readdir state. */ +/* There are two structures here, rdstate which is a hash table + * of the second structure which is a filldir_node. */ + +/* This is a struct kmem_cache for filldir nodes, because we allocate a lot of them + * and they shouldn't waste memory. If the node has a small name (as defined + * by the dentry structure), then we use an inline name to preserve kmalloc + * space. */ +static struct kmem_cache *unionfs_filldir_cachep; +int init_filldir_cache(void) +{ + unionfs_filldir_cachep = + kmem_cache_create("unionfs_filldir", sizeof(struct filldir_node), 0, + SLAB_RECLAIM_ACCOUNT, NULL); + + if (!unionfs_filldir_cachep) + return -ENOMEM; + + return 0; +} + +void destroy_filldir_cache(void) +{ + if (!unionfs_filldir_cachep) + return; + kmem_cache_destroy(unionfs_filldir_cachep); + return; +} + +/* This is a tuning parameter that tells us roughly how big to make the + * hash table in directory entries per page. This isn't perfect, but + * at least we get a hash table size that shouldn't be too overloaded. + * The following averages are based on my home directory. + * 14.44693 Overall + * 12.29 Single Page Directories + * 117.93 Multi-page directories + */ +#define DENTPAGE 4096 +#define DENTPERONEPAGE 12 +#define DENTPERPAGE 118 +#define MINHASHSIZE 1 +static int guesstimate_hash_size(struct inode *inode) +{ + struct inode *hidden_inode; + int bindex; + int hashsize = MINHASHSIZE; + + if (itopd(inode)->uii_hashsize > 0) + return itopd(inode)->uii_hashsize; + + for (bindex = ibstart(inode); bindex <= ibend(inode); bindex++) { + if (!(hidden_inode = itohi_index(inode, bindex))) + continue; + + if (hidden_inode->i_size == DENTPAGE) { + hashsize += DENTPERONEPAGE; + } else { + hashsize += + (hidden_inode->i_size / DENTPAGE) * DENTPERPAGE; + } + } + + return hashsize; +} + +int init_rdstate(struct file *file) +{ + BUG_ON(sizeof(loff_t) != (sizeof(unsigned int) + sizeof(unsigned int))); + BUG_ON(ftopd(file)->rdstate != NULL); + + ftopd(file)->rdstate = + alloc_rdstate(file->f_dentry->d_inode, fbstart(file)); + if (!ftopd(file)->rdstate) + return -ENOMEM; + return 0; +} + +struct unionfs_dir_state *find_rdstate(struct inode *inode, loff_t fpos) +{ + struct unionfs_dir_state *rdstate = NULL; + struct list_head *pos; + + print_entry("f_pos: %lld", fpos); + spin_lock(&itopd(inode)->uii_rdlock); + list_for_each(pos, &itopd(inode)->uii_readdircache) { + struct unionfs_dir_state *r = + list_entry(pos, struct unionfs_dir_state, uds_cache); + if (fpos == rdstate2offset(r)) { + itopd(inode)->uii_rdcount--; + list_del(&r->uds_cache); + rdstate = r; + break; + } + } + spin_unlock(&itopd(inode)->uii_rdlock); + print_exit_pointer(rdstate); + return rdstate; +} + +struct unionfs_dir_state *alloc_rdstate(struct inode *inode, int bindex) +{ + int i = 0; + int hashsize; + int mallocsize = sizeof(struct unionfs_dir_state); + struct unionfs_dir_state *rdstate; + + hashsize = guesstimate_hash_size(inode); + mallocsize += hashsize * sizeof(struct list_head); + /* Round it up to the next highest power of two. */ + mallocsize--; + mallocsize |= mallocsize >> 1; + mallocsize |= mallocsize >> 2; + mallocsize |= mallocsize >> 4; + mallocsize |= mallocsize >> 8; + mallocsize |= mallocsize >> 16; + mallocsize++; + + /* This should give us about 500 entries anyway. */ + if (mallocsize > PAGE_SIZE) + mallocsize = PAGE_SIZE; + + hashsize = + (mallocsize - + sizeof(struct unionfs_dir_state)) / sizeof(struct list_head); + + rdstate = KMALLOC(mallocsize, GFP_KERNEL); + if (!rdstate) + return NULL; + + spin_lock(&itopd(inode)->uii_rdlock); + if (itopd(inode)->uii_cookie >= (MAXRDCOOKIE - 1)) + itopd(inode)->uii_cookie = 1; + else + itopd(inode)->uii_cookie++; + + rdstate->uds_cookie = itopd(inode)->uii_cookie; + spin_unlock(&itopd(inode)->uii_rdlock); + rdstate->uds_offset = 1; + rdstate->uds_access = jiffies; + rdstate->uds_bindex = bindex; + rdstate->uds_dirpos = 0; + rdstate->uds_hashentries = 0; + rdstate->uds_size = hashsize; + for (i = 0; i < rdstate->uds_size; i++) + INIT_LIST_HEAD(&rdstate->uds_list[i]); + + return rdstate; +} + +static void free_filldir_node(struct filldir_node *node) +{ + if (node->namelen >= DNAME_INLINE_LEN_MIN) + KFREE(node->name); + kmem_cache_free(unionfs_filldir_cachep, node); +} + +void free_rdstate(struct unionfs_dir_state *state) +{ + struct filldir_node *tmp; + int i; + + for (i = 0; i < state->uds_size; i++) { + struct list_head *head = &(state->uds_list[i]); + struct list_head *pos, *n; + + /* traverse the list and deallocate space */ + list_for_each_safe(pos, n, head) { + tmp = list_entry(pos, struct filldir_node, file_list); + list_del(&tmp->file_list); + free_filldir_node(tmp); + } + } + + KFREE(state); +} + +struct filldir_node *find_filldir_node(struct unionfs_dir_state *rdstate, + const char *name, int namelen) +{ + int index; + unsigned int hash; + struct list_head *head; + struct list_head *pos; + struct filldir_node *cursor = NULL; + int found = 0; + + /* If we print entry, we end up with spurious data. */ + /* print_entry("name = %*s", namelen, name); */ + print_entry_location(); + + BUG_ON(namelen <= 0); + + hash = full_name_hash(name, namelen); + index = hash % rdstate->uds_size; + + head = &(rdstate->uds_list[index]); + list_for_each(pos, head) { + cursor = list_entry(pos, struct filldir_node, file_list); + + if (cursor->namelen == namelen && cursor->hash == hash + && !strncmp(cursor->name, name, namelen)) { + /* a duplicate exists, and hence no need to create entry to the list */ + found = 1; + /* if the duplicate is in this branch, then the file system is corrupted. */ + if (cursor->bindex == rdstate->uds_bindex) { + //buf->error = err = -EIO; + dprint(PRINT_DEBUG, + "Possible I/O error unionfs_filldir: a file is duplicated in the same branch %d: %s\n", + rdstate->uds_bindex, cursor->name); + } + break; + } + } + + if (!found) { + cursor = NULL; + } + print_exit_pointer(cursor); + return cursor; +} + +inline struct filldir_node *alloc_filldir_node(const char *name, int namelen, + unsigned int hash, int bindex) +{ + struct filldir_node *newnode; + + newnode = + (struct filldir_node *)kmem_cache_alloc(unionfs_filldir_cachep, + GFP_KERNEL); + if (!newnode) + goto out; + + out: + return newnode; +} + +int add_filldir_node(struct unionfs_dir_state *rdstate, const char *name, + int namelen, int bindex, int whiteout) +{ + struct filldir_node *new; + unsigned int hash; + int index; + int err = 0; + struct list_head *head; + + /* We can't print this because we end up Oopsing. */ + /* print_entry("name = %*s", namelen, name); */ + print_entry_location(); + + BUG_ON(namelen <= 0); + + hash = full_name_hash(name, namelen); + index = hash % rdstate->uds_size; + head = &(rdstate->uds_list[index]); + + new = alloc_filldir_node(name, namelen, hash, bindex); + if (!new) { + err = -ENOMEM; + goto out; + } + + INIT_LIST_HEAD(&new->file_list); + new->namelen = namelen; + new->hash = hash; + new->bindex = bindex; + new->whiteout = whiteout; + + if (namelen < DNAME_INLINE_LEN_MIN) { + new->name = new->iname; + } else { + new->name = (char *)KMALLOC(namelen + 1, GFP_KERNEL); + if (!new->name) { + kmem_cache_free(unionfs_filldir_cachep, new); + new = NULL; + goto out; + } + } + + memcpy(new->name, name, namelen); + new->name[namelen] = '\0'; + + rdstate->uds_hashentries++; + + list_add(&(new->file_list), head); + out: + print_exit_status(err); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/rename.c @@ -0,0 +1,943 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: rename.c,v 1.47 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +static int do_rename(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry, + int bindex, struct dentry **wh_old) +{ + int err = 0; + struct dentry *hidden_old_dentry; + struct dentry *hidden_new_dentry; + struct dentry *hidden_old_dir_dentry; + struct dentry *hidden_new_dir_dentry; + struct dentry *hidden_wh_dentry; + struct dentry *hidden_wh_dir_dentry; + char *wh_name = NULL; + + print_entry(" bindex=%d", bindex); + + print_dentry("IN: do_rename, old_dentry", old_dentry); + print_dentry("IN: do_rename, new_dentry", new_dentry); + dprint(PRINT_DEBUG, "do_rename for bindex = %d\n", bindex); + + hidden_new_dentry = dtohd_index(new_dentry, bindex); + hidden_old_dentry = dtohd_index(old_dentry, bindex); + + if (!hidden_new_dentry) { + hidden_new_dentry = + create_parents(new_dentry->d_parent->d_inode, new_dentry, + bindex); + if (IS_ERR(hidden_new_dentry)) { + dprint(PRINT_DEBUG, + "error creating directory tree for rename, bindex = %d\n", + bindex); + err = PTR_ERR(hidden_new_dentry); + goto out; + } + } + + wh_name = alloc_whname(new_dentry->d_name.name, new_dentry->d_name.len); + if (IS_ERR(wh_name)) { + err = PTR_ERR(wh_name); + goto out; + } + + hidden_wh_dentry = + LOOKUP_ONE_LEN(wh_name, hidden_new_dentry->d_parent, + new_dentry->d_name.len + WHLEN); + if (IS_ERR(hidden_wh_dentry)) { + err = PTR_ERR(hidden_wh_dentry); + goto out; + } + + if (hidden_wh_dentry->d_inode) { + /* get rid of the whiteout that is existing */ + if (hidden_new_dentry->d_inode) { + printk(KERN_WARNING + "Both a whiteout and a dentry exist when doing a rename!\n"); + err = -EIO; + + DPUT(hidden_wh_dentry); + goto out; + } + + hidden_wh_dir_dentry = lock_parent(hidden_wh_dentry); + if (!(err = is_robranch_super(old_dentry->d_sb, bindex))) { + err = + vfs_unlink(hidden_wh_dir_dentry->d_inode, + hidden_wh_dentry, NULL); + } + DPUT(hidden_wh_dentry); + unlock_dir(hidden_wh_dir_dentry); + if (err) + goto out; + } else + DPUT(hidden_wh_dentry); + + DGET(hidden_old_dentry); + hidden_old_dir_dentry = GET_PARENT(hidden_old_dentry); + hidden_new_dir_dentry = GET_PARENT(hidden_new_dentry); + + lock_rename(hidden_old_dir_dentry, hidden_new_dir_dentry); + + err = is_robranch_super(old_dentry->d_sb, bindex); + if (err) + goto out_unlock; + + /* ready to whiteout for old_dentry. + caller will create the actual whiteout, + and must dput(*wh_old) */ + if (wh_old) { + char *whname; + whname = alloc_whname(old_dentry->d_name.name, + old_dentry->d_name.len); + err = PTR_ERR(whname); + if (IS_ERR(whname)) + goto out_unlock; + *wh_old = LOOKUP_ONE_LEN(whname, hidden_old_dir_dentry, + old_dentry->d_name.len + WHLEN); + KFREE(whname); + err = PTR_ERR(*wh_old); + if (IS_ERR(*wh_old)) { + *wh_old = NULL; + goto out_unlock; + } + } + + print_dentry("NEWBEF", new_dentry); + print_dentry("OLDBEF", old_dentry); + err = vfs_rename(hidden_old_dir_dentry->d_inode, hidden_old_dentry, + NULL, hidden_new_dir_dentry->d_inode, + hidden_new_dentry, NULL); + print_dentry("NEWAFT", new_dentry); + print_dentry("OLDAFT", old_dentry); + + out_unlock: + unlock_rename(hidden_old_dir_dentry, hidden_new_dir_dentry); + + DPUT(hidden_old_dir_dentry); + DPUT(hidden_new_dir_dentry); + DPUT(hidden_old_dentry); + + out: + if (!err) { + /* Fixup the newdentry. */ + if (bindex < dbstart(new_dentry)) + set_dbstart(new_dentry, bindex); + else if (bindex > dbend(new_dentry)) + set_dbend(new_dentry, bindex); + } + + KFREE(wh_name); + + print_dentry("OUT: do_rename, old_dentry", old_dentry); + print_dentry("OUT: do_rename, new_dentry", new_dentry); + + print_exit_status(err); + return err; +} + +static int unionfs_rename_whiteout(struct inode *old_dir, + struct dentry *old_dentry, + struct inode *new_dir, + struct dentry *new_dentry) +{ + int err = 0; + int bindex, bwh_old; + int old_bstart, old_bend; + int new_bstart, new_bend; + int do_copyup = -1; + struct dentry *parent_dentry; + int local_err = 0; + int eio = 0; + int revert = 0; + struct dentry *wh_old = NULL; + + print_entry_location(); + + old_bstart = dbstart(old_dentry); + bwh_old = old_bstart; + old_bend = dbend(old_dentry); + parent_dentry = old_dentry->d_parent; + + new_bstart = dbstart(new_dentry); + new_bend = dbend(new_dentry); + + /* Rename source to destination. */ + err = do_rename(old_dir, old_dentry, new_dir, new_dentry, old_bstart, + &wh_old); + if (err) { + if (!IS_COPYUP_ERR(err)) { + goto out; + } + do_copyup = old_bstart - 1; + } else { + revert = 1; + } + + /* Unlink all instances of destination that exist to the left of + * bstart of source. On error, revert back, goto out. + */ + for (bindex = old_bstart - 1; bindex >= new_bstart; bindex--) { + struct dentry *unlink_dentry; + struct dentry *unlink_dir_dentry; + + unlink_dentry = dtohd_index(new_dentry, bindex); + if (!unlink_dentry) { + continue; + } + + unlink_dir_dentry = lock_parent(unlink_dentry); + if (!(err = is_robranch_super(old_dir->i_sb, bindex))) { + err = + vfs_unlink(unlink_dir_dentry->d_inode, + unlink_dentry, NULL); + } + + fist_copy_attr_times(new_dentry->d_parent->d_inode, + unlink_dir_dentry->d_inode); + /* propagate number of hard-links */ + new_dentry->d_parent->d_inode->i_nlink = + get_nlinks(new_dentry->d_parent->d_inode); + + unlock_dir(unlink_dir_dentry); + if (!err) { + if (bindex != new_bstart) { + DPUT(unlink_dentry); + set_dtohd_index(new_dentry, bindex, NULL); + } + } else if (IS_COPYUP_ERR(err)) { + do_copyup = bindex - 1; + } else if (revert) { + DPUT(wh_old); + goto revert; + } + } + + if (do_copyup != -1) { + for (bindex = do_copyup; bindex >= 0; bindex--) { + /* copyup the file into some left directory, so that you can rename it */ + err = + copyup_dentry(old_dentry->d_parent->d_inode, + old_dentry, old_bstart, bindex, NULL, + old_dentry->d_inode->i_size); + if (!err) { + DPUT(wh_old); + bwh_old = bindex; + err = + do_rename(old_dir, old_dentry, new_dir, + new_dentry, bindex, &wh_old); + break; + } + } + } + + /* make it opaque */ + if (S_ISDIR(old_dentry->d_inode->i_mode)) { + err = make_dir_opaque(old_dentry, dbstart(old_dentry)); + if (err) + goto revert; + } + + /* Create whiteout for source, only if: + * (1) There is more than one underlying instance of source. + * (2) We did a copy_up + */ + if ((old_bstart != old_bend) || (do_copyup != -1)) { + struct dentry *hidden_parent; + BUG_ON(!wh_old || IS_ERR(wh_old) || wh_old->d_inode + || bwh_old < 0); + hidden_parent = lock_parent(wh_old); + local_err = vfs_create(hidden_parent->d_inode, wh_old, S_IRUGO, + NULL); + unlock_dir(hidden_parent); + if (!local_err) + set_dbopaque(old_dentry, bwh_old); + else { + /* We can't fix anything now, so we cop-out and use -EIO. */ + printk + ("<0>We can't create a whiteout for the source in rename!\n"); + err = -EIO; + } + } + + out: + DPUT(wh_old); + print_exit_status(err); + return err; + + revert: + /* Do revert here. */ + local_err = unionfs_refresh_hidden_dentry(new_dentry, old_bstart); + if (local_err) { + printk(KERN_WARNING + "Revert failed in rename: the new refresh failed.\n"); + eio = -EIO; + } + + local_err = unionfs_refresh_hidden_dentry(old_dentry, old_bstart); + if (local_err) { + printk(KERN_WARNING + "Revert failed in rename: the old refresh failed.\n"); + eio = -EIO; + goto revert_out; + } + + if (!dtohd_index(new_dentry, bindex) + || !dtohd_index(new_dentry, bindex)->d_inode) { + printk(KERN_WARNING + "Revert failed in rename: the object disappeared from under us!\n"); + eio = -EIO; + goto revert_out; + } + + if (dtohd_index(old_dentry, bindex) + && dtohd_index(old_dentry, bindex)->d_inode) { + printk(KERN_WARNING + "Revert failed in rename: the object was created underneath us!\n"); + eio = -EIO; + goto revert_out; + } + + local_err = + do_rename(new_dir, new_dentry, old_dir, old_dentry, old_bstart, + NULL); + + /* If we can't fix it, then we cop-out with -EIO. */ + if (local_err) { + printk(KERN_WARNING "Revert failed in rename!\n"); + eio = -EIO; + } + + local_err = unionfs_refresh_hidden_dentry(new_dentry, bindex); + if (local_err) + eio = -EIO; + local_err = unionfs_refresh_hidden_dentry(old_dentry, bindex); + if (local_err) + eio = -EIO; + + revert_out: + if (eio) + err = eio; + print_exit_status(err); + return err; +} + +/* + * Unfortunately, we cannot simply call things like dbstart() in different + * places of the rename code because we move things around. So, we use this + * structure to pass the necessary information around to all the places that + * need it. + */ +struct rename_info { + int do_copyup; + int do_whiteout; + int rename_ok; + + int old_bstart; + int old_bend; + int new_bstart; + int new_bend; + + int isdir; /* Is the source a directory? */ + int clobber; /* Are we clobbering the destination? */ + + int bwh_old; /* where we create the whiteout */ + struct dentry *wh_old; /* lookup and set by do_rename() */ +}; +#ifdef UNIONFS_DELETE_ALL +/* + * Rename all occurences of source except for the leftmost destination + */ +static int __rename_all(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry, + fd_set * success_mask, struct rename_info *info) +{ + int bindex; + int err = 0; + + print_entry_location(); + + /* Loop through all the branches from right to left and rename all + * instances of source to destination, except the leftmost destination + */ + for (bindex = info->old_bend; bindex >= info->old_bstart; bindex--) { + /* We don't rename if there is no source. */ + if (dtohd_index(old_dentry, bindex) == NULL) + continue; + + /* we rename the bstart of destination only at the last of + * all operations, so that we don't lose it on error + */ + if (info->clobber && (bindex == info->new_bstart)) + continue; + + DPUT(info->wh_old); + info->bwh_old = bindex; + /* We shouldn't have a handle on this if there is no inode. */ + err = + do_rename(old_dir, old_dentry, new_dir, new_dentry, bindex, + &info->wh_old); + if (!err) { + /* For reverting. */ + FD_SET(bindex, success_mask); + /* So we know not to copyup on failures the right */ + info->rename_ok = bindex; + } else if (IS_COPYUP_ERR(err)) { + if (info->isdir) { + err = -EXDEV; + break; + } + + /* we need a whiteout... */ + info->do_whiteout = bindex - 1; + + if (bindex == info->old_bstart) + /* ...and a copyup */ + info->do_copyup = bindex - 1; + + err = 0; /* reset error */ + } else + break; /* error is set by do_rename */ + } + + print_exit_status(err); + return err; +} + +/* + * Unlink all destinations (if they exist) to the left of the left-most + * source + */ +static int __rename_all_unlink(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry, + struct rename_info *info) +{ + int bindex; + + struct dentry *unlink_dentry; + struct dentry *unlink_dir_dentry; + + int err = 0; + + print_entry_location(); + + for (bindex = info->old_bstart - 1; bindex > info->new_bstart; bindex--) { + unlink_dentry = dtohd_index(new_dentry, bindex); + if (!unlink_dentry) + continue; + + /* lock, unlink if possible, copyup times, unlock */ + unlink_dir_dentry = lock_parent(unlink_dentry); + if (!(err = is_robranch_super(old_dir->i_sb, bindex))) + err = + vfs_unlink(unlink_dir_dentry->d_inode, + unlink_dentry, NULL); + + fist_copy_attr_times(new_dentry->d_parent->d_inode, + unlink_dir_dentry->d_inode); + new_dentry->d_parent->d_inode->i_nlink = + get_nlinks(new_dentry->d_parent->d_inode); + + unlock_dir(unlink_dir_dentry); + + if (!err) { + if (bindex != info->new_bstart) { + DPUT(unlink_dentry); + set_dtohd_index(new_dentry, bindex, NULL); + } + } else if (IS_COPYUP_ERR(err)) { + if (info->isdir) { + err = -EXDEV; + break; + } + info->do_copyup = bindex - 1; + + err = 0; /* reset error */ + } else + break; /* err is set by is_ro_branch_super or vfs_unlink */ + } + + print_exit_status(err); + return err; +} + +/* + * Try to revert everything we have done in __rename_all and __rename_all_unlink + */ +static int __rename_all_revert(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry, + fd_set * success_mask, struct rename_info *info) +{ + int bindex; + + int err; + int eio = 0; + + print_entry_location(); + + for (bindex = info->old_bstart; bindex <= info->old_bend; bindex++) { + if (!FD_ISSET(bindex, success_mask)) + continue; + + err = unionfs_refresh_hidden_dentry(new_dentry, bindex); + if (err) { + printk(KERN_WARNING "Revert failed in rename: " + "the new refresh failed.\n"); + eio = -EIO; + } + + err = unionfs_refresh_hidden_dentry(old_dentry, bindex); + if (err) { + printk(KERN_WARNING "Revert failed in rename: " + "the old refresh failed.\n"); + eio = -EIO; + continue; + } + + if (!dtohd_index(new_dentry, bindex) + || !dtohd_index(new_dentry, bindex)->d_inode) { + printk(KERN_WARNING "Revert failed in rename: " + "the object disappeared from under us!\n"); + eio = -EIO; + continue; + } + + if (dtohd_index(old_dentry, bindex) + && dtohd_index(old_dentry, bindex)->d_inode) { + printk(KERN_WARNING "Revert failed in rename: " + "the object was created underneath us!\n"); + eio = -EIO; + continue; + } + + err = + do_rename(new_dir, new_dentry, old_dir, old_dentry, bindex, + NULL); + /* If we can't fix it, then we cop-out with -EIO. */ + if (err) { + printk(KERN_WARNING "Revert failed in rename!\n"); + eio = -EIO; + } + + err = unionfs_refresh_hidden_dentry(new_dentry, bindex); + if (err) + eio = -EIO; + err = unionfs_refresh_hidden_dentry(old_dentry, bindex); + if (err) + eio = -EIO; + } + + print_exit_status(eio); + return eio; +} + +/* + * Finish off the rename, by either over writing the last destination or + * unlinking the last destination to the left of us + */ +static int __rename_all_clobber(struct inode *old_dir, + struct dentry *old_dentry, + struct inode *new_dir, + struct dentry *new_dentry, + struct rename_info *info) +{ + int err = 0; + + print_entry_location(); + + if (dtohd_index(old_dentry, info->new_bstart)) { + /* rename the last source, knowing we're overwriting something */ + DPUT(info->wh_old); + info->bwh_old = info->new_bstart; + err = + do_rename(old_dir, old_dentry, new_dir, new_dentry, + info->new_bstart, &info->wh_old); + if (IS_COPYUP_ERR(err)) { + if (info->isdir) { + err = -EXDEV; + goto out; + } + if (info->rename_ok > info->new_bstart) { + if ((info->do_copyup == -1) + || (info->new_bstart - 1 < info->do_copyup)) + info->do_copyup = info->new_bstart - 1; + } + if ((info->do_whiteout == -1) + || (info->new_bstart - 1 < info->do_whiteout)) { + info->do_whiteout = info->new_bstart - 1; + } + err = 0; // reset error + } + } else if (info->new_bstart < info->old_bstart) { + /* the newly renamed file would get hidden, let's unlink the + * file to the left of it */ + struct dentry *unlink_dentry; + struct dentry *unlink_dir_dentry; + + unlink_dentry = dtohd_index(new_dentry, info->new_bstart); + + unlink_dir_dentry = lock_parent(unlink_dentry); + if (!(err = is_robranch_super(old_dir->i_sb, info->new_bstart))) + err = vfs_unlink(unlink_dir_dentry->d_inode, + unlink_dentry, NULL); + + fist_copy_attr_times(new_dentry->d_parent->d_inode, + unlink_dir_dentry->d_inode); + new_dentry->d_parent->d_inode->i_nlink = + get_nlinks(new_dentry->d_parent->d_inode); + + unlock_dir(unlink_dir_dentry); + + if (IS_COPYUP_ERR(err)) { + if (info->isdir) { + err = -EXDEV; + goto out; + } + if ((info->do_copyup == -1) + || (info->new_bstart - 1 < info->do_copyup)) + info->do_copyup = info->new_bstart - 1; + + err = 0; // reset error + } + } + + out: + print_exit_status(err); + return err; +} + +/* + * The function is nasty, nasty, nasty, but so is rename. :( + */ +static int unionfs_rename_all(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry) +{ + struct dentry *parent_dentry = NULL; + int err = 0; + int eio; + + /* These variables control error handling. */ + fd_set success_mask; + char *name = NULL; + + /* unfortunately, we have to resort to this, because dbstart/dbend would + return different things in different place of the rename code */ + struct rename_info info; + + info.rename_ok = FD_SETSIZE; /* The last rename that is ok. */ + info.do_copyup = -1; /* Where we should start copyup. */ + info.do_whiteout = -1; /* Where we should start whiteouts of the source. */ + info.wh_old = NULL; + info.bwh_old = -1; + + print_entry_location(); + + parent_dentry = old_dentry->d_parent; + name = KMALLOC(old_dentry->d_name.len + 1, GFP_KERNEL); + if (!name) { + err = -ENOMEM; + goto out; + } + strncpy(name, old_dentry->d_name.name, old_dentry->d_name.len + 1); + + info.new_bstart = dbstart(new_dentry); + info.new_bend = dbend(new_dentry); + + info.old_bstart = dbstart(old_dentry); + info.old_bend = dbend(old_dentry); + + BUG_ON(info.new_bstart < 0); + BUG_ON(info.old_bstart < 0); + + /* The failure mask only can deal with FD_SETSIZE entries. */ + BUG_ON(info.old_bend > FD_SETSIZE); + BUG_ON(info.new_bend > FD_SETSIZE); + FD_ZERO(&success_mask); + + /* Life is simpler if the dentry doesn't exist. */ + info.clobber = + (dtohd_index(new_dentry, info.new_bstart)->d_inode) ? 1 : 0; + info.isdir = S_ISDIR(old_dentry->d_inode->i_mode); + + /* rename everything we can */ + err = + __rename_all(old_dir, old_dentry, new_dir, new_dentry, + &success_mask, &info); + if (err) + goto revert; + + /* unlink destinations even further left */ + err = + __rename_all_unlink(old_dir, old_dentry, new_dir, new_dentry, + &info); + if (err) + goto revert; + + if (info.clobber) { + /* Now we need to handle the leftmost of the destination. */ + err = + __rename_all_clobber(old_dir, old_dentry, new_dir, + new_dentry, &info); + if (err) + goto revert; + } + + /* Copy up if necessary */ + if (info.do_copyup != -1) { + int bindex; + + for (bindex = info.do_copyup; bindex >= 0; bindex--) { + err = + copyup_dentry(old_dentry->d_parent->d_inode, + old_dentry, info.old_bstart, bindex, + NULL, old_dentry->d_inode->i_size); + if (!err) { + DPUT(info.wh_old); + info.bwh_old = bindex; + err = + do_rename(old_dir, old_dentry, new_dir, + new_dentry, bindex, &info.wh_old); + break; + } + } + } + + /* make it opaque */ + if (S_ISDIR(old_dentry->d_inode->i_mode)) { + err = make_dir_opaque(old_dentry, dbstart(old_dentry)); + if (err) + goto revert; + } + + /* Create a whiteout for the source. */ + if (info.do_whiteout != -1) { + struct dentry *hidden_parent; + BUG_ON(info.do_whiteout < 0 + || !info.wh_old || IS_ERR(info.wh_old) + || info.wh_old->d_inode || info.bwh_old < 0); + hidden_parent = lock_parent(info.wh_old); + err = vfs_create(hidden_parent->d_inode, info.wh_old, S_IRUGO, + NULL); + unlock_dir(hidden_parent); + if (!err) + set_dbopaque(old_dentry, info.bwh_old); + else { + /* We can't fix anything now, so we -EIO. */ + printk(KERN_WARNING "We can't create a whiteout for the" + "source in rename!\n"); + err = -EIO; + goto out; + } + } + + /* We are at the point where reverting doesn't happen. */ + goto out; + + revert: + /* something bad happened, try to revert */ + eio = + __rename_all_revert(old_dir, old_dentry, new_dir, new_dentry, + &success_mask, &info); + if (eio) + err = eio; + + out: + DPUT(info.wh_old); + KFREE(name); + print_exit_status(err); + return err; +} +#endif + +static struct dentry *lookup_whiteout(struct dentry *dentry) +{ + char *whname; + int bindex = -1, bstart = -1, bend = -1; + struct dentry *parent, *hidden_parent, *wh_dentry; + + whname = alloc_whname(dentry->d_name.name, dentry->d_name.len); + if (IS_ERR(whname)) + return (void *)whname; + + parent = GET_PARENT(dentry); + lock_dentry(parent); + bstart = dbstart(parent); + bend = dbend(parent); + wh_dentry = ERR_PTR(-ENOENT); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_parent = dtohd_index(parent, bindex); + if (!hidden_parent) + continue; + wh_dentry = + LOOKUP_ONE_LEN(whname, hidden_parent, + dentry->d_name.len + WHLEN); + if (IS_ERR(wh_dentry)) + continue; + if (wh_dentry->d_inode) + break; + DPUT(wh_dentry); + wh_dentry = ERR_PTR(-ENOENT); + } + unlock_dentry(parent); + DPUT(parent); + KFREE(whname); + return wh_dentry; +} + +/* We can't copyup a directory, because it may involve huge + * numbers of children, etc. Doing that in the kernel would + * be bad, so instead we let the userspace recurse and ask us + * to copy up each file separately + */ +static int may_rename_dir(struct dentry *dentry) +{ + int err, bstart; + + err = check_empty(dentry, NULL); + if (err == -ENOTEMPTY) { + if (is_robranch(dentry)) + return -EXDEV; + } else if (err) + return err; + + bstart = dbstart(dentry); + if (dbend(dentry) == bstart || dbopaque(dentry) == bstart) + return 0; + + set_dbstart(dentry, bstart + 1); + err = check_empty(dentry, NULL); + set_dbstart(dentry, bstart); + if (err == -ENOTEMPTY) + err = -EXDEV; + return err; +} + +int unionfs_rename(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry) +{ + int err = 0; + struct dentry *wh_dentry; + + print_entry_location(); + + double_lock_dentry(old_dentry, new_dentry); + + checkinode(old_dir, "unionfs_rename-old_dir"); + checkinode(new_dir, "unionfs_rename-new_dir"); + print_dentry("IN: unionfs_rename, old_dentry", old_dentry); + print_dentry("IN: unionfs_rename, new_dentry", new_dentry); + + if (!S_ISDIR(old_dentry->d_inode->i_mode)) + err = unionfs_partial_lookup(old_dentry); + else + err = may_rename_dir(old_dentry); + + if (err) + goto out; + + err = unionfs_partial_lookup(new_dentry); + if (err) + goto out; + + /* + * if new_dentry is already hidden because of whiteout, + * simply override it even if the whiteouted dir is not empty. + */ + wh_dentry = lookup_whiteout(new_dentry); + if (!IS_ERR(wh_dentry)) + DPUT(wh_dentry); + else if (new_dentry->d_inode) { + if (S_ISDIR(old_dentry->d_inode->i_mode) != + S_ISDIR(new_dentry->d_inode->i_mode)) { + err = + S_ISDIR(old_dentry->d_inode-> + i_mode) ? -ENOTDIR : -EISDIR; + goto out; + } + + if (S_ISDIR(new_dentry->d_inode->i_mode)) { + struct unionfs_dir_state *namelist; + /* check if this unionfs directory is empty or not */ + err = check_empty(new_dentry, &namelist); + if (err) + goto out; + + if (!is_robranch(new_dentry)) + err = delete_whiteouts(new_dentry, + dbstart(new_dentry), + namelist); + + free_rdstate(namelist); + + if (err) + goto out; + } + } +#ifdef UNIONFS_DELETE_ALL + if (IS_SET(old_dir->i_sb, DELETE_ALL)) + err = unionfs_rename_all(old_dir, old_dentry, new_dir, + new_dentry); + else +#endif + err = unionfs_rename_whiteout(old_dir, old_dentry, new_dir, + new_dentry); + + out: + checkinode(new_dir, "post unionfs_rename-new_dir"); + print_dentry("OUT: unionfs_rename, old_dentry", old_dentry); + + if (err) { + /* clear the new_dentry stuff created */ + d_drop(new_dentry); + } else { + /* force re-lookup since the dir on ro branch is not renamed, + and hidden dentries still indicate the un-renamed ones. */ + if (S_ISDIR(old_dentry->d_inode->i_mode)) + atomic_dec(&dtopd(old_dentry)->udi_generation); + print_dentry("OUT: unionfs_rename, new_dentry", + new_dentry); + } + + unlock_dentry(new_dentry); + unlock_dentry(old_dentry); + print_exit_status(err); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/sioq.c @@ -0,0 +1,109 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2004-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ + +#include "unionfs.h" + +struct workqueue_struct *sioq; + +int __init init_sioq(void) +{ + int err; + + sioq = create_workqueue("unionfs_siod"); + if (!IS_ERR(sioq)) + return 0; + + err = PTR_ERR(sioq); + printk(KERN_ERR "create_workqueue failed %d\n", err); + sioq = NULL; + return err; +} + +void fin_sioq(void) +{ + if (sioq) + destroy_workqueue(sioq); +} + +void run_sioq(work_func_t func, struct sioq_args *args) +{ + INIT_WORK(&args->wk, func); + + init_completion(&args->comp); + while (!queue_work(sioq, &args->wk)) { + // TODO: do accounting if needed + schedule(); + } + wait_for_completion(&args->comp); +} + +void __unionfs_create(struct work_struct *work) +{ + struct sioq_args *args = container_of(work, struct sioq_args, wk); + struct create_args *c = &args->create; + args->err = vfs_create(c->parent, c->dentry, c->mode, c->nd); + complete(&args->comp); +} + +void __unionfs_mkdir(struct work_struct *work) +{ + struct sioq_args *args = container_of(work, struct sioq_args, wk); + struct mkdir_args *m = &args->mkdir; + args->err = vfs_mkdir(m->parent, m->dentry, NULL, m->mode); + complete(&args->comp); +} + +void __unionfs_mknod(struct work_struct *work) +{ + struct sioq_args *args = container_of(work, struct sioq_args, wk); + struct mknod_args *m = &args->mknod; + args->err = vfs_mknod(m->parent, m->dentry, NULL, m->mode, m->dev); + complete(&args->comp); +} +void __unionfs_symlink(struct work_struct *work) +{ + struct sioq_args *args = container_of(work, struct sioq_args, wk); + struct symlink_args *s = &args->symlink; + args->err = vfs_symlink(s->parent, s->dentry, NULL, s->symbuf); + complete(&args->comp); +} + +void __unionfs_unlink(struct work_struct *work) +{ + struct sioq_args *args = container_of(work, struct sioq_args, wk); + struct unlink_args *u = &args->unlink; + args->err = vfs_unlink(u->parent, u->dentry, NULL); + complete(&args->comp); +} + +void __delete_whiteouts(struct work_struct *work) { + struct sioq_args *args = container_of(work, struct sioq_args, wk); + struct deletewh_args *d = &args->deletewh; + args->err = delete_whiteouts(d->dentry, d->bindex, d->namelist); + complete(&args->comp); +} + +void __is_opaque_dir(struct work_struct *work) +{ + struct sioq_args *args = container_of(work, struct sioq_args, wk); + + args->ret = lookup_one_len(UNIONFS_DIR_OPAQUE, args->isopaque.dentry, + sizeof(UNIONFS_DIR_OPAQUE) - 1); + complete(&args->comp); +} --- /dev/null +++ b/fs/unionfs/sioq.h @@ -0,0 +1,80 @@ +#ifndef _SIOQ_H +#define _SIOQ_H + +struct deletewh_args { + struct unionfs_dir_state *namelist; + struct dentry *dentry; + int bindex; +}; + +struct isopaque_args { + struct dentry *dentry; +}; + +struct create_args { + struct inode *parent; + struct dentry *dentry; + umode_t mode; + struct nameidata *nd; +}; + +struct mkdir_args { + struct inode *parent; + struct dentry *dentry; + umode_t mode; +}; + +struct mknod_args { + struct inode *parent; + struct dentry *dentry; + umode_t mode; + dev_t dev; +}; + +struct symlink_args { + struct inode *parent; + struct dentry *dentry; + char *symbuf; + umode_t mode; +}; + +struct unlink_args { + struct inode *parent; + struct dentry *dentry; +}; + + +struct sioq_args { + + struct completion comp; + struct work_struct wk; + int err; + void *ret; + + union { + struct deletewh_args deletewh; + struct isopaque_args isopaque; + struct create_args create; + struct mkdir_args mkdir; + struct mknod_args mknod; + struct symlink_args symlink; + struct unlink_args unlink; + }; //} u; +}; + +extern struct workqueue_struct *sioq; +int __init init_sioq(void); +extern void fin_sioq(void); +extern void run_sioq(work_func_t func, struct sioq_args *args); + +/* Extern definitions for our privledge escalation helpers */ +extern void __unionfs_create(struct work_struct *work); +extern void __unionfs_mkdir(struct work_struct *work); +extern void __unionfs_mknod(struct work_struct *work); +extern void __unionfs_symlink(struct work_struct *work); +extern void __unionfs_unlink(struct work_struct *work); +extern void __delete_whiteouts(struct work_struct *work); +extern void __is_opaque_dir(struct work_struct *work); + +#endif /* _SIOQ_H */ + --- /dev/null +++ b/fs/unionfs/stale_inode.c @@ -0,0 +1,133 @@ +/* + * Adpated from linux/fs/bad_inode.c + * + * Copyright (C) 1997, Stephen Tweedie + * + * Provide stub functions for "stale" inodes, a bit friendlier than the + * -EIO that bad_inode.c does. + */ +/* + * $Id: stale_inode.c,v 1.13 2006/03/21 09:22:11 jsipek Exp $ + */ + +#include <linux/version.h> + +#include <linux/fs.h> +#include <linux/stat.h> +#include <linux/sched.h> + +static struct address_space_operations unionfs_stale_aops; + +/* declarations for "sparse */ +extern struct inode_operations stale_inode_ops; + +/* + * The follow_link operation is special: it must behave as a no-op + * so that a stale root inode can at least be unmounted. To do this + * we must dput() the base and return the dentry with a dget(). + */ +static void *stale_follow_link(struct dentry *dent, struct nameidata *nd) +{ + int err = vfs_follow_link(nd, ERR_PTR(-ESTALE)); + return ERR_PTR(err); +} + +static int return_ESTALE(void) +{ + return -ESTALE; +} + +#define ESTALE_ERROR ((void *) (return_ESTALE)) + +static struct file_operations stale_file_ops = { + .llseek = ESTALE_ERROR, + .read = ESTALE_ERROR, + .write = ESTALE_ERROR, + .readdir = ESTALE_ERROR, + .poll = ESTALE_ERROR, + .ioctl = ESTALE_ERROR, + .mmap = ESTALE_ERROR, + .open = ESTALE_ERROR, + .flush = ESTALE_ERROR, + .release = ESTALE_ERROR, + .fsync = ESTALE_ERROR, + .fasync = ESTALE_ERROR, + .lock = ESTALE_ERROR, +}; + +struct inode_operations stale_inode_ops = { + .create = ESTALE_ERROR, + .lookup = ESTALE_ERROR, + .link = ESTALE_ERROR, + .unlink = ESTALE_ERROR, + .symlink = ESTALE_ERROR, + .mkdir = ESTALE_ERROR, + .rmdir = ESTALE_ERROR, + .mknod = ESTALE_ERROR, + .rename = ESTALE_ERROR, + .readlink = ESTALE_ERROR, + .follow_link = stale_follow_link, + .truncate = ESTALE_ERROR, + .permission = ESTALE_ERROR, +}; + +/* + * When a filesystem is unable to read an inode due to an I/O error in + * its read_inode() function, it can call make_stale_inode() to return a + * set of stubs which will return ESTALE errors as required. + * + * We only need to do limited initialisation: all other fields are + * preinitialised to zero automatically. + */ + +/** + * make_stale_inode - mark an inode stale due to an I/O error + * @inode: Inode to mark stale + * + * When an inode cannot be read due to a media or remote network + * failure this function makes the inode "stale" and causes I/O operations + * on it to fail from this point on. + */ + +void make_stale_inode(struct inode *inode) +{ + inode->i_mode = S_IFREG; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + inode->i_op = &stale_inode_ops; + inode->i_fop = &stale_file_ops; + inode->i_mapping->a_ops = &unionfs_stale_aops; +} + +/* + * This tests whether an inode has been flagged as stale. The test uses + * &stale_inode_ops to cover the case of invalidated inodes as well as + * those created by make_stale_inode() above. + */ + +/** + * is_stale_inode - is an inode errored + * @inode: inode to test + * + * Returns true if the inode in question has been marked as stale. + */ + +int is_stale_inode(struct inode *inode) +{ + return (inode->i_op == &stale_inode_ops); +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/subr.c @@ -0,0 +1,204 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: subr.c,v 1.142 2006/09/21 18:19:36 jsipek Exp $ + */ + +#include "unionfs.h" +#include <linux/security.h> + +/* Pass an unionfs dentry and an index. It will try to create a whiteout + * for the filename in dentry, and will try in branch 'index'. On error, + * it will proceed to a branch to the left. + */ +int create_whiteout(struct dentry *dentry, int start) +{ + int bstart, bend, bindex; + struct dentry *hidden_dir_dentry; + struct dentry *hidden_dentry; + struct dentry *hidden_wh_dentry; + char *name = NULL; + int err = -EINVAL; + + print_entry("start = %d", start); + + verify_locked(dentry); + + print_dentry("IN create_whiteout", dentry); + bstart = dbstart(dentry); + bend = dbend(dentry); + + /* create dentry's whiteout equivalent */ + name = alloc_whname(dentry->d_name.name, dentry->d_name.len); + if (IS_ERR(name)) { + err = PTR_ERR(name); + goto out; + } + + for (bindex = start; bindex >= 0; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + + if (!hidden_dentry) { + /* if hidden dentry is not present, create the entire + * hidden dentry directory structure and go ahead. + * Since we want to just create whiteout, we only want + * the parent dentry, and hence get rid of this dentry. + */ + hidden_dentry = create_parents(dentry->d_inode, + dentry, bindex); + if (!hidden_dentry || IS_ERR(hidden_dentry)) { + dprint(PRINT_DEBUG_WHITEOUT, + "create_parents failed for bindex = %d\n", + bindex); + continue; + } + } + hidden_wh_dentry = + LOOKUP_ONE_LEN(name, hidden_dentry->d_parent, + dentry->d_name.len + WHLEN); + if (IS_ERR(hidden_wh_dentry)) + continue; + + /* The whiteout already exists. This used to be impossible, but + * now is possible because of opaqueness. */ + if (hidden_wh_dentry->d_inode) { + DPUT(hidden_wh_dentry); + err = 0; + goto out; + } + + hidden_dir_dentry = lock_parent(hidden_wh_dentry); + if (!(err = is_robranch_super(dentry->d_sb, bindex))) { + err = + vfs_create(hidden_dir_dentry->d_inode, + hidden_wh_dentry, + ~current->fs->umask & S_IRWXUGO, NULL); + + } + unlock_dir(hidden_dir_dentry); + DPUT(hidden_wh_dentry); + + if (!err) + break; + + if (!IS_COPYUP_ERR(err)) + break; + } + + /* set dbopaque so that lookup will not proceed after this branch */ + if (!err) + set_dbopaque(dentry, bindex); + + print_dentry("OUT create_whiteout", dentry); + out: + KFREE(name); + print_exit_status(err); + return err; +} + +/* This is a helper function for rename, which ends up with hosed over dentries + * when it needs to revert. */ +int unionfs_refresh_hidden_dentry(struct dentry *dentry, int bindex) +{ + struct dentry *hidden_dentry; + struct dentry *hidden_parent; + int err = 0; + + print_entry(" bindex = %d", bindex); + + verify_locked(dentry); + lock_dentry(dentry->d_parent); + hidden_parent = dtohd_index(dentry->d_parent, bindex); + unlock_dentry(dentry->d_parent); + + BUG_ON(!S_ISDIR(hidden_parent->d_inode->i_mode)); + + hidden_dentry = + LOOKUP_ONE_LEN(dentry->d_name.name, hidden_parent, + dentry->d_name.len); + if (IS_ERR(hidden_dentry)) { + err = PTR_ERR(hidden_dentry); + goto out; + } + + if (dtohd_index(dentry, bindex)) + DPUT(dtohd_index(dentry, bindex)); + if (itohi_index(dentry->d_inode, bindex)) { + IPUT(itohi_index(dentry->d_inode, bindex)); + set_itohi_index(dentry->d_inode, bindex, NULL); + } + if (!hidden_dentry->d_inode) { + DPUT(hidden_dentry); + set_dtohd_index(dentry, bindex, NULL); + } else { + set_dtohd_index(dentry, bindex, hidden_dentry); + set_itohi_index(dentry->d_inode, bindex, + IGRAB(hidden_dentry->d_inode)); + } + + out: + print_exit_status(err); + return err; +} + +int make_dir_opaque(struct dentry *dentry, int bindex) +{ + int err; + struct dentry *hidden_dentry, *diropq; + struct inode *hidden_dir; + + hidden_dentry = dtohd_index(dentry, bindex); + hidden_dir = hidden_dentry->d_inode; + BUG_ON(!S_ISDIR(dentry->d_inode->i_mode) + || !S_ISDIR(hidden_dir->i_mode)); + + mutex_lock(&hidden_dir->i_mutex); + diropq = LOOKUP_ONE_LEN(UNIONFS_DIR_OPAQUE, hidden_dentry, + sizeof(UNIONFS_DIR_OPAQUE) - 1); + err = PTR_ERR(diropq); + if (IS_ERR(diropq)) + goto out; + err = 0; + + if (!diropq->d_inode) + err = vfs_create(hidden_dir, diropq, S_IRUGO, NULL); + DPUT(diropq); + if (!err) + set_dbopaque(dentry, bindex); + + out: + mutex_unlock(&hidden_dir->i_mutex); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/super.c @@ -0,0 +1,762 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: super.c,v 1.101 2006/11/04 22:27:51 jsipek Exp $ + */ + +#include "unionfs.h" + +/* The inode cache is used with alloc_inode for both our inode info and the + * vfs inode. */ +static struct kmem_cache *unionfs_inode_cachep; + +void unionfs_read_inode(struct inode *inode) +{ +#ifdef UNIONFS_MMAP + /* SP: use real address operations */ + extern struct address_space_operations unionfs_aops; +#else + static struct address_space_operations unionfs_empty_aops; +#endif + int size; + + print_entry_location(); + + if (!itopd(inode)) { + printk(KERN_ERR + "No kernel memory when allocating inode private data!\n"); + BUG(); + } + + memset(itopd(inode), 0, sizeof(struct unionfs_inode_info)); + itopd(inode)->b_start = -1; + itopd(inode)->b_end = -1; + atomic_set(&itopd(inode)->uii_generation, + atomic_read(&stopd(inode->i_sb)->usi_generation)); + itopd(inode)->uii_rdlock = SPIN_LOCK_UNLOCKED; + itopd(inode)->uii_rdcount = 1; + itopd(inode)->uii_hashsize = -1; + INIT_LIST_HEAD(&itopd(inode)->uii_readdircache); + + size = sbmax(inode->i_sb) * sizeof(struct inode *); + itohi_ptr(inode) = KZALLOC(size, GFP_KERNEL); + if (!itohi_ptr(inode)) { + printk(KERN_ERR + "No kernel memory when allocating lower-pointer array!\n"); + BUG(); + } + + inode->i_version++; + inode->i_op = &unionfs_main_iops; + inode->i_fop = &unionfs_main_fops; +#ifdef UNIONFS_MMAP + inode->i_mapping->a_ops = &unionfs_aops; +#else + /* I don't think ->a_ops is ever allowed to be NULL */ + inode->i_mapping->a_ops = &unionfs_empty_aops; + dprint(PRINT_DEBUG, "setting inode 0x%p a_ops to empty (0x%p)\n", + inode, inode->i_mapping->a_ops); +#endif + + print_exit_location(); +} + +#if 0 +static void unionfs_put_inode(struct inode *inode) +{ + print_entry_location(); + + dprint(PRINT_DEBUG, "%s i_count = %d, i_nlink = %d\n", __FUNCTION__, + atomic_read(&inode->i_count), inode->i_nlink); + + /* + * This is really funky stuff: + * Basically, if i_count == 1, iput will then decrement it and this + * inode will be destroyed. It is currently holding a reference to the + * hidden inode. Therefore, it needs to release that reference by + * calling iput on the hidden inode. iput() _will_ do it for us (by + * calling our clear_inode), but _only_ if i_nlink == 0. The problem + * is, NFS keeps i_nlink == 1 for silly_rename'd files. So we must for + * our i_nlink to 0 here to trick iput() into calling our clear_inode. + */ + + if (atomic_read(&inode->i_count) == 1) + inode->i_nlink = 0; + + print_exit_location(); +} +#endif + +/* + * we now define delete_inode, because there are two VFS paths that may + * destroy an inode: one of them calls clear inode before doing everything + * else that's needed, and the other is fine. This way we truncate the inode + * size (and its pages) and then clear our own inode, which will do an iput + * on our and the lower inode. + */ +static void unionfs_delete_inode(struct inode *inode) +{ + print_entry_location(); + + checkinode(inode, "unionfs_delete_inode IN"); + inode->i_size = 0; /* every f/s seems to do that */ + +#ifdef UNIONFS_MMAP + /* SP: if you try to clear_inode() when + * inode->i_data.nrpages != 0, you'll hit a BUG + * this is also what generic_delete_inode does */ + if (inode->i_data.nrpages) + truncate_inode_pages(&inode->i_data, 0); +#endif + clear_inode(inode); + + print_exit_location(); +} + +/* final actions when unmounting a file system */ +static void unionfs_put_super(struct super_block *sb) +{ + int bindex, bstart, bend; + struct unionfs_sb_info *spd; + + print_entry_location(); + + if ((spd = stopd(sb))) { +#ifdef UNIONFS_IMAP + /* XXX: Free persistent inode stuff. */ + cleanup_imap_data(sb); +#endif + bstart = sbstart(sb); + bend = sbend(sb); + for (bindex = bstart; bindex <= bend; bindex++) + mntput(stohiddenmnt_index(sb, bindex)); + + /* Make sure we have no leaks of branchget/branchput. */ + for (bindex = bstart; bindex <= bend; bindex++) + BUG_ON(branch_count(sb, bindex) != 0); + + KFREE(spd->usi_data); + KFREE(spd); + stopd_lhs(sb) = NULL; + } + dprint(PRINT_DEBUG, "unionfs: released super\n"); + + print_exit_location(); +} + +/* Since people use this to answer the "How big of a file can I write?" + * question, we report the size of the highest priority branch as the size of + * the union. + */ +static int unionfs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + int err = 0; + struct super_block *sb, *hidden_sb; + + sb = dentry->d_sb; + + hidden_sb = stohs_index(sb, sbstart(sb)); + err = vfs_statfs(hidden_sb->s_root, buf); + + buf->f_type = UNIONFS_SUPER_MAGIC; + buf->f_namelen -= WHLEN; + + memset(&buf->f_fsid, 0, sizeof(__kernel_fsid_t)); + memset(&buf->f_spare, 0, sizeof(buf->f_spare)); + + return err; +} + +static int do_binary_remount(struct super_block *sb, int *flags, char *data) +{ + unsigned long *uldata = (unsigned long *)data; + int err; + + uldata++; + + switch (*uldata) { + case UNIONFS_IOCTL_DELBRANCH: + err = unionfs_ioctl_delbranch(sb, *(uldata + 1)); + break; + default: + err = -ENOTTY; + } + + return err; +} + +/* We don't support a standard text remount, but we do have a magic remount + * for unionctl. The idea is that you can remove a branch without opening + * the union. Eventually it would be nice to support a full-on remount, so + * that you can have all of the directories change at once, but that would + * require some pretty complicated matching code. */ +static int unionfs_remount_fs(struct super_block *sb, int *flags, char *data) +{ + if (data && *((unsigned long *)data) == UNIONFS_REMOUNT_MAGIC) + return do_binary_remount(sb, flags, data); + printk("Warning! dirs delete and imap options to remount are ignored\n"); + return 0; +} + +/* + * Called by iput() when the inode reference count reached zero + * and the inode is not hashed anywhere. Used to clear anything + * that needs to be, before the inode is completely destroyed and put + * on the inode free list. + */ +static void unionfs_clear_inode(struct inode *inode) +{ + int bindex, bstart, bend; + struct inode *hidden_inode; + struct list_head *pos, *n; + struct unionfs_dir_state *rdstate; + + print_entry_location(); + + checkinode(inode, "unionfs_clear_inode IN"); + + list_for_each_safe(pos, n, &itopd(inode)->uii_readdircache) { + rdstate = list_entry(pos, struct unionfs_dir_state, uds_cache); + list_del(&rdstate->uds_cache); + free_rdstate(rdstate); + } + + /* Decrement a reference to a hidden_inode, which was incremented + * by our read_inode when it was created initially. */ + bstart = ibstart(inode); + bend = ibend(inode); + if (bstart >= 0) { + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_inode = itohi_index(inode, bindex); + if (!hidden_inode) + continue; + IPUT(hidden_inode); + } + } + // XXX: why this assertion fails? + // because it doesn't like us + // BUG_ON((inode->i_state & I_DIRTY) != 0); + KFREE(itohi_ptr(inode)); + itohi_ptr(inode) = NULL; + + print_exit_location(); +} + +static struct inode *unionfs_alloc_inode(struct super_block *sb) +{ + struct unionfs_inode_container *c; + + print_entry_location(); + + c = (struct unionfs_inode_container *) + kmem_cache_alloc(unionfs_inode_cachep, GFP_KERNEL); + if (!c) { + print_exit_pointer(NULL); + return NULL; + } + + memset(&c->info, 0, sizeof(c->info)); + + c->vfs_inode.i_version = 1; + print_exit_pointer(&c->vfs_inode); + return &c->vfs_inode; +} + +static void unionfs_destroy_inode(struct inode *inode) +{ + print_entry("inode = %p", inode); + kmem_cache_free(unionfs_inode_cachep, itopd(inode)); + print_exit_location(); +} + +static void init_once(void *v) +{ + struct unionfs_inode_container *c = (struct unionfs_inode_container *)v; + + print_entry_location(); + + inode_init_once(&c->vfs_inode); + + print_exit_location(); +} + +int init_inode_cache(void) +{ + int err = 0; + + print_entry_location(); + + unionfs_inode_cachep = + kmem_cache_create("unionfs_inode_cache", + sizeof(struct unionfs_inode_container), 0, + SLAB_RECLAIM_ACCOUNT, init_once); + if (!unionfs_inode_cachep) + err = -ENOMEM; + print_exit_status(err); + return err; +} + +void destroy_inode_cache(void) +{ + print_entry_location(); + if (!unionfs_inode_cachep) + goto out; + kmem_cache_destroy(unionfs_inode_cachep); + out: + print_exit_location(); + return; +} + +/* Called when we have a dirty inode, right here we only throw out + * parts of our readdir list that are too old. + */ +static int unionfs_write_inode(struct inode *inode, int sync) +{ + struct list_head *pos, *n; + struct unionfs_dir_state *rdstate; + + print_entry_location(); + + spin_lock(&itopd(inode)->uii_rdlock); + list_for_each_safe(pos, n, &itopd(inode)->uii_readdircache) { + rdstate = list_entry(pos, struct unionfs_dir_state, uds_cache); + /* We keep this list in LRU order. */ + if ((rdstate->uds_access + RDCACHE_JIFFIES) > jiffies) + break; + itopd(inode)->uii_rdcount--; + list_del(&rdstate->uds_cache); + free_rdstate(rdstate); + } + spin_unlock(&itopd(inode)->uii_rdlock); + + print_exit_location(); + return 0; +} + +/* + * Used only in nfs, to kill any pending RPC tasks, so that subsequent + * code can actually succeed and won't leave tasks that need handling. + * + * PS. I wonder if this is somehow useful to undo damage that was + * left in the kernel after a user level file server (such as amd) + * dies. + */ +static void unionfs_umount_begin(struct super_block *sb) +{ + struct super_block *hidden_sb; + int bindex, bstart, bend; + + print_entry_location(); +#if 0 + if (!(flags & MNT_FORCE)) + /* we are not being MNT_FORCEd, therefore we should emulate old + * behaviour + */ + goto out; +#endif + bstart = sbstart(sb); + bend = sbend(sb); + for (bindex = bstart; bindex <= bend; bindex++) { + hidden_sb = stohs_index(sb, bindex); + + if (hidden_sb && hidden_sb->s_op && + hidden_sb->s_op->umount_begin) + hidden_sb->s_op->umount_begin(hidden_sb); + } + + print_exit_location(); +} + +static int unionfs_show_options(struct seq_file *m, struct vfsmount *mnt) +{ + struct super_block *sb = mnt->mnt_sb; + int ret = 0; + unsigned long tmp = 0; + char *hidden_path; + int bindex, bstart, bend; + int perms; + + lock_dentry(sb->s_root); + + tmp = __get_free_page(GFP_KERNEL); + if (!tmp) { + ret = -ENOMEM; + goto out; + } + + bindex = bstart = sbstart(sb); + bend = sbend(sb); + + seq_printf(m, ",dirs="); + for (bindex = bstart; bindex <= bend; bindex++) { + struct path tp; + tp.dentry = dtohd_index(sb->s_root, bindex); + tp.mnt = stohiddenmnt_index(sb, bindex); + + hidden_path = + d_path(&tp, (char *)tmp, PAGE_SIZE); + perms = branchperms(sb, bindex); + seq_printf(m, "%s=%s", hidden_path, + perms & MAY_WRITE ? "rw" : + perms & MAY_NFSRO ? "nfsro" : "ro"); + if (bindex != bend) { + seq_printf(m, ":"); + } + } + + seq_printf(m, ",debug=%u", get_debug_mask()); + +#ifdef UNIONFS_DELETE_ALL + if (IS_SET(sb, DELETE_ALL)) + seq_printf(m, ",delete=all"); + else +#endif + seq_printf(m, ",delete=whiteout"); + out: + if (tmp) + free_page(tmp); + unlock_dentry(sb->s_root); + return ret; +} + +#ifdef CONFIG_EXPORTFS +/* + * export operations. + * unionfs cannot handle disconnected dentry, since it has no hidden dentries. + */ +/* un-tested 64 bit environment (pointer and inode number) */ + +#define is_anon(d) ((d)->d_flags & DCACHE_DISCONNECTED) +extern struct export_operations export_op_default; + +static void prepend_path(char **path, const char *name, int len) +{ + *path -= len; + memcpy(*path, name, len); + (*path)--; + **path = '/'; +} + +struct filldir_arg { + int found, called; + char *path; + ino_t ino, parent_ino; +}; + +static int filldir(void *arg, const char *name, int len, loff_t pos, ino_t ino, + unsigned int d_type) +{ + struct filldir_arg *a = arg; + + a->called++; + if (len == 2 && !strncmp(name, "..", 2)) { + a->parent_ino = ino; + a->found++; + } else if (ino == a->ino) { + if (len != 1 || *name != '.') + prepend_path(&a->path, name, len); + a->found++; + } + return (a->found == 2) ? 1 : 0; +} + +static struct dentry *get_hidden_parent(struct super_block *hidden_sb, + ino_t hidden_parent_ino) +{ + __u32 fh[2]; + + if (hidden_sb->s_root->d_inode->i_ino == hidden_parent_ino) + return DGET(hidden_sb->s_root); + + fh[0] = hidden_parent_ino; + fh[1] = 0; + return export_op_default.get_dentry(hidden_sb, fh); +} + +static struct dentry *do_get_dentry(struct super_block *sb, ino_t ino, + __u32 gen, struct dentry *hidden_root, + ino_t hidden_ino, ino_t hidden_parent_ino) +{ + struct dentry *dentry, *hidden_parent, *parent; + char *path, *p; + struct filldir_arg arg = { + .ino = hidden_ino, + .parent_ino = hidden_parent_ino + }; + int open_flags, err, bindex, bend, found; + struct file *hidden_file; + struct super_block *hidden_sb; + + print_entry("hr%p, hi%lu, hpi%lu", + hidden_root, hidden_ino, hidden_parent_ino); + + dentry = ERR_PTR(-ENOMEM); + path = __getname(); + if (!path) + goto out; + arg.path = path + PATH_MAX - 1; + *arg.path = 0; + + open_flags = O_RDONLY | O_DIRECTORY /* | O_NOATIME */ ; + if (force_o_largefile()) + open_flags |= O_LARGEFILE; + + dentry = ERR_PTR(-ESTALE); + unionfs_read_lock(sb); + lock_dentry(sb->s_root); + bend = dbend(sb->s_root); + found = -1; + for (bindex = 0; found == -1 && bindex <= bend; bindex++) + if (hidden_root == dtohd_index(sb->s_root, bindex)) + found = bindex; + unlock_dentry(sb->s_root); + if (found == -1) + goto out_unlock; + + bindex = found; + hidden_sb = stohs_index(sb, bindex); + while (1) { + hidden_parent = get_hidden_parent(hidden_sb, hidden_parent_ino); + dentry = hidden_parent; + if (IS_ERR(hidden_parent)) + goto out_unlock; + + branchget(sb, bindex); + hidden_file = DENTRY_OPEN(DGET(hidden_parent), NULL, + open_flags); + if (IS_ERR(hidden_file)) { + dentry = (void *)hidden_file; + DPUT(hidden_parent); + branchput(sb, bindex); + goto out_unlock; + } + + arg.found = 0; + while (arg.found != 2) { + arg.called = 0; + err = vfs_readdir(hidden_file, filldir, &arg); + if (!arg.called || err < 0) + break; + } + fput(hidden_file); + branchput(sb, bindex); + if (arg.found != 2) { + dentry = ERR_PTR(-ESTALE); + DPUT(hidden_parent); + goto out_unlock; + } + + DPUT(hidden_parent); + if (hidden_parent_ino == hidden_root->d_inode->i_ino) + break; + arg.ino = hidden_parent_ino; + hidden_parent_ino = arg.parent_ino; + } + BUG_ON(arg.path < path); + + parent = DGET(sb->s_root); + p = strchr(++arg.path, '/'); + while (p) { + mutex_lock(&parent->d_inode->i_mutex); + dentry = LOOKUP_ONE_LEN(arg.path, parent, p - arg.path); + mutex_unlock(&parent->d_inode->i_mutex); + DPUT(parent); + if (IS_ERR(dentry)) + goto out_unlock; + if (!dentry->d_inode || !S_ISDIR(dentry->d_inode->i_mode)) { + DPUT(dentry); + dentry = ERR_PTR(-ESTALE); + goto out_unlock; + } + parent = dentry; + arg.path = p + 1; + p = strchr(arg.path, '/'); + } + mutex_lock(&parent->d_inode->i_mutex); + dentry = LOOKUP_ONE_LEN(arg.path, parent, strlen(arg.path)); + mutex_unlock(&parent->d_inode->i_mutex); + DPUT(parent); + if (!IS_ERR(dentry) + && (!dentry->d_inode + || dentry->d_inode->i_ino != ino + || dentry->d_inode->i_generation != gen)) { + DPUT(dentry); + dentry = ERR_PTR(-ESTALE); + } + + out_unlock: + unionfs_read_unlock(sb); + __putname(path); + out: + print_exit_pointer(dentry); + return dentry; +} + +enum { + FhHead = 4, FhHRoot1 = FhHead, FhHRoot2, + FhHIno1, FhHIno2, FhHPIno1, FhHPIno2, + FhTail +}; + +static void do_decode(__u32 * fh, struct dentry **hidden_root, + ino_t * hidden_ino, ino_t * hidden_parent_ino) +{ + unsigned long root; + + root = fh[FhHRoot2]; + *hidden_ino = fh[FhHIno2]; + *hidden_parent_ino = fh[FhHPIno2]; +#if BITS_PER_LONG == 64 + root |= ((unsigned long)fh[FhHRoot1]) << 32; + *hidden_ino |= ((unsigned long) fh[FhHIno1]) << 32; + *hidden_parent_ino |= ((unsigned long) fh[FhHPIno1]) << 32; +#elif BITS_PER_LONG == 32 + /* ok */ +#else +#error unknown size +#endif + + *hidden_root = (struct dentry*) root; +} + +static int unionfs_encode_fh(struct dentry *dentry, __u32 * fh, int *max_len, + int connectable) +{ + int type, len, bindex; + struct super_block *sb; + struct dentry *h_root; + ino_t h_ino, hp_ino; + static int warn; + + print_entry("dentry %p", dentry); + BUG_ON(is_anon(dentry) || !dentry->d_inode + || is_anon(dentry->d_parent)); + +#ifdef UNIONFS_IMAP + if (!warn && stopd(dentry->d_sb)->usi_persistent) + warn++; +#endif + if (!warn) { + printk(KERN_WARNING "Exporting Unionfs without imap" + " option may stop your NFS server or client"); + warn++; + } + + sb = dentry->d_sb; + unionfs_read_lock(sb); + lock_dentry(dentry); + + len = *max_len; + type = export_op_default.encode_fh(dentry, fh, max_len, connectable); + if (type == 255 || *max_len > FhHead || len < FhTail) { + type = 255; + goto out; + } + + *max_len = FhTail; + bindex = dbstart(dentry); + lock_dentry(sb->s_root); + h_root = dtohd_index(sb->s_root, bindex); + unlock_dentry(sb->s_root); + h_ino = itohi_index(dentry->d_inode, bindex)->i_ino; + hp_ino = parent_ino(dtohd(dentry)); + fh[FhHRoot2] = (unsigned long) h_root; + fh[FhHIno2] = h_ino; + fh[FhHPIno2] = hp_ino; +#if BITS_PER_LONG == 64 + fh[FhHRoot1] = ((unsigned long) h_root) >> 32; + fh[FhHIno1] = h_ino >> 32; + fh[FhHPIno1] = hp_ino >> 32; +#endif + + out: + unionfs_print(PRINT_MAIN_EXIT, "%d, fh{i%u, g%d, hr%x, hi%u, hpi%u}\n", + type, fh[0], fh[1], fh[FhHRoot2], fh[FhHIno2], + fh[FhHPIno2]); + unlock_dentry(dentry); + unionfs_read_unlock(sb); + return type; +} + +static struct dentry *unionfs_decode_fh(struct super_block *sb, __u32 * fh, + int fh_len, int fh_type, + int (*acceptable) (void *context, + struct dentry * de), + void *context) +{ + struct dentry *dentry, *hidden_root; + ino_t hidden_ino, hidden_parent_ino; + + print_entry("%d, fh{i%u, g%d, hr%x, hi%u, hpi%u}", + fh_type, fh[0], fh[1], fh[FhHRoot2], fh[FhHIno2], + fh[FhHPIno2]); + + dentry = export_op_default.get_dentry(sb, fh); + if (!dentry || IS_ERR(dentry) || (dentry->d_inode && !is_anon(dentry))) + return dentry; + + d_drop(dentry); + DPUT(dentry); + do_decode(fh, &hidden_root, &hidden_ino, &hidden_parent_ino); + dentry = do_get_dentry(sb, fh[0], fh[1], hidden_root, hidden_ino, + hidden_parent_ino); + if (!IS_ERR(dentry)) { + if (acceptable(context, dentry)) + return dentry; /* success */ + DPUT(dentry); + dentry = NULL; + } + return dentry; +} + +struct export_operations unionfs_export_ops = { + .decode_fh = unionfs_decode_fh, + .encode_fh = unionfs_encode_fh +}; +#endif + +struct super_operations unionfs_sops = { + //.put_inode = unionfs_put_inode, + .delete_inode = unionfs_delete_inode, + .put_super = unionfs_put_super, + .statfs = unionfs_statfs, + .remount_fs = unionfs_remount_fs, + .clear_inode = unionfs_clear_inode, + .umount_begin = unionfs_umount_begin, + .show_options = unionfs_show_options, + .write_inode = unionfs_write_inode, + .alloc_inode = unionfs_alloc_inode, + .destroy_inode = unionfs_destroy_inode, +}; + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/unionfs.h @@ -0,0 +1,736 @@ +#ifndef __UNIONFS_H_ +#define __UNIONFS_H_ + +#ifdef __KERNEL__ + +#include <linux/version.h> +#include <linux/sched.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/string.h> +#include <linux/stat.h> +#include <linux/errno.h> +#include <linux/wait.h> +#include <linux/limits.h> +#include <linux/random.h> +#include <linux/poll.h> +#include <linux/buffer_head.h> +#include <linux/pagemap.h> +#include <linux/namei.h> +#include <linux/module.h> +#include <linux/mount.h> +#include <linux/page-flags.h> +#include <linux/writeback.h> +#include <linux/page-flags.h> +#include <linux/statfs.h> +#include <linux/smp.h> +#include <linux/smp_lock.h> +#include <linux/file.h> +#include <linux/slab.h> +#include <linux/vmalloc.h> +#include <linux/poll.h> +#include <linux/list.h> +#include <linux/init.h> +#include <linux/xattr.h> +#include <linux/security.h> +#include <linux/spinlock.h> +#include <linux/compat.h> + +#include <linux/swap.h> + +#include <asm/system.h> +#include <asm/mman.h> +#include <linux/seq_file.h> +#include <linux/dcache.h> +#include <linux/poll.h> + +#ifndef UNIONFS_UNSUPPORTED +#if LINUX_VERSION_CODE != KERNEL_VERSION(SUP_MAJOR,SUP_MINOR,SUP_PATCH) +#warning You are compiling Unionfs on an unsupported kernel version. +#warning To compile Unionfs, you will need to define UNIONFS_UNSUPPORTED. +#warning Try adding: EXTRACFLAGS=-DUNIONFS_UNSUPPORTED to fistdev.mk +#endif +#endif + +/* the file system name */ +#define UNIONFS_NAME "unionfs" + +/* unionfs file systems superblock magic */ +#define UNIONFS_SUPER_MAGIC 0xf15f083d + +/* unionfs root inode number */ +#define UNIONFS_ROOT_INO 1 + +/* Mount time flags */ +#define MOUNT_FLAG(sb) (stopd(sb)->usi_mount_flag) + +/* number of characters while generating unique temporary file names */ +#define UNIONFS_TMPNAM_LEN 12 + +/* Operations vectors defined in specific files. */ +extern struct file_operations unionfs_main_fops; +extern struct file_operations unionfs_dir_fops; +extern struct inode_operations unionfs_main_iops; +extern struct inode_operations unionfs_dir_iops; +extern struct inode_operations unionfs_symlink_iops; +extern struct super_operations unionfs_sops; +extern struct dentry_operations unionfs_dops; +#ifdef CONFIG_EXPORTFS +extern struct export_operations unionfs_export_ops; +#endif + +/* How long should an entry be allowed to persist */ +#define RDCACHE_JIFFIES 5*HZ + +/* file private data. */ +struct unionfs_file_info { + int b_start; + int b_end; + atomic_t ufi_generation; + + struct unionfs_dir_state *rdstate; + struct file **ufi_file; +}; + +/* unionfs inode data in memory */ +struct unionfs_inode_info { + int b_start; + int b_end; + atomic_t uii_generation; + int uii_stale; + /* Stuff for readdir over NFS. */ + spinlock_t uii_rdlock; + struct list_head uii_readdircache; + int uii_rdcount; + int uii_hashsize; + int uii_cookie; + /* The hidden inodes */ + struct inode **uii_inode; + /* to keep track of reads/writes for unlinks before closes */ + atomic_t uii_totalopens; +}; + +struct unionfs_inode_container { + struct unionfs_inode_info info; + struct inode vfs_inode; +}; + +/* unionfs dentry data in memory */ +struct unionfs_dentry_info { + /* The semaphore is used to lock the dentry as soon as we get into a + * unionfs function from the VFS. Our lock ordering is that children + * go before their parents. */ + struct semaphore udi_sem; + int udi_bstart; + int udi_bend; + int udi_bopaque; + int udi_bcount; + atomic_t udi_generation; + struct dentry **udi_dentry; +}; + +/* A putmap is used so that older files can still do branchput correctly. */ +struct putmap { + atomic_t count; + int bend; + int map[0]; +}; + +/* These are the pointers to our various objects. */ +struct unionfs_usi_data { + struct super_block *sb; + struct vfsmount *hidden_mnt; + atomic_t sbcount; + int branchperms; +}; + +/* unionfs super-block data in memory */ +struct unionfs_sb_info { + int b_end; + + atomic_t usi_generation; + unsigned long usi_mount_flag; + struct rw_semaphore usi_rwsem; + + struct unionfs_usi_data *usi_data; + + /* These map branch numbers for old generation numbers to the new bindex, + * so that branchput will behave properly. */ + int usi_firstputmap; + int usi_lastputmap; + struct putmap **usi_putmaps; + +#ifdef UNIONFS_IMAP + int usi_persistent; + /* These will need a lock. */ + uint64_t usi_next_avail; + uint8_t usi_num_bmapents; + struct bmapent *usi_bmap; + struct file *usi_forwardmap; + struct file **usi_reversemaps; + struct file **usi_map_table; + int *usi_bnum_table; //This is a table of branches to fsnums. +#endif /* UNIONFS_IMAP */ +}; + +/* + * structure for making the linked list of entries by readdir on left branch + * to compare with entries on right branch + */ +struct filldir_node { + struct list_head file_list; // list for directory entries + char *name; // name entry + int hash; // name hash + int namelen; // name len since name is not 0 terminated + int bindex; // we can check for duplicate whiteouts and files in the same branch in order to return -EIO. + int whiteout; // is this a whiteout entry? + char iname[DNAME_INLINE_LEN_MIN]; // Inline name, so we don't need to separately kmalloc small ones +}; + +/* Directory hash table. */ +struct unionfs_dir_state { + unsigned int uds_cookie; /* The cookie, which is based off of uii_rdversion */ + unsigned int uds_offset; /* The entry we have returned. */ + int uds_bindex; + loff_t uds_dirpos; /* The offset within the lower level directory. */ + int uds_size; /* How big is the hash table? */ + int uds_hashentries; /* How many entries have been inserted? */ + unsigned long uds_access; + /* This cache list is used when the inode keeps us around. */ + struct list_head uds_cache; + struct list_head uds_list[0]; +}; + +/* privileged io workqueue */ +#include "sioq.h" + +/* include miscellaneous macros */ +#include "unionfs_macros.h" + +/* include debug macros */ +#include "unionfs_debug.h" + +/* include persistent imap code */ +#include "unionfs_imap.h" + +/* Cache creation/deletion routines. */ +void destroy_filldir_cache(void); +int init_filldir_cache(void); +int init_inode_cache(void); +void destroy_inode_cache(void); +int init_dentry_cache(void); +void destroy_dentry_cache(void); + +/* Initialize and free readdir-specific state. */ +int init_rdstate(struct file *file); +struct unionfs_dir_state *alloc_rdstate(struct inode *inode, int bindex); +struct unionfs_dir_state *find_rdstate(struct inode *inode, loff_t fpos); +void free_rdstate(struct unionfs_dir_state *state); +int add_filldir_node(struct unionfs_dir_state *rdstate, const char *name, + int namelen, int bindex, int whiteout); +struct filldir_node *find_filldir_node(struct unionfs_dir_state *rdstate, + const char *name, int namelen); + +struct dentry **alloc_new_dentries(int objs); +struct unionfs_usi_data *alloc_new_data(int objs); + +#ifdef FIST_MALLOC_DEBUG + +extern void *unionfs_kzalloc(size_t size, gfp_t flags, int line, + const char *file); +extern void *unionfs_kmalloc(size_t size, gfp_t flags, int line, + const char *file); +extern void unionfs_kfree(void *ptr, int line, const char *file); + +extern struct dentry *unionfs_dget_parent(struct dentry *child, int line, + const char *file); +extern struct dentry *unionfs_dget(struct dentry *ptr, int line, + const char *file); +extern void unionfs_dput(struct dentry *ptr, int line, const char *file); +extern struct inode *unionfs_igrab(struct inode *inode, int line, char *file); +extern void unionfs_iput(struct inode *inode, int line, char *file); +extern struct dentry *unionfs_lookup_one_len(const char *name, + struct dentry *parent, int len, + int line, const char *file); +void record_path_lookup(struct nameidata *nd, int line, const char *file); +void record_path_release(struct nameidata *nd, int line, const char *file); +struct file *unionfs_dentry_open(struct dentry *ptr, struct vfsmount *mnt, + int flags, int line, const char *file); +void record_set(struct dentry *upper, int index, struct dentry *ptr, + struct dentry *old, int line, const char *file); + +#define KZALLOC(size,flags) unionfs_kzalloc((size),(flags),__LINE__,__FILE__) +#define KMALLOC(size,flags) unionfs_kmalloc((size),(flags),__LINE__,__FILE__) +#define KFREE(ptr) unionfs_kfree((ptr),__LINE__,__FILE__) +#define DGET(d) unionfs_dget((d),__LINE__,__FILE__) +#define DPUT(d) unionfs_dput((d),__LINE__,__FILE__) +# define IPUT(a) unionfs_iput((a),__LINE__,__FILE__) +# define IGRAB(a) unionfs_igrab((a),__LINE__,__FILE__) +#define LOOKUP_ONE_LEN(name,parent,len) unionfs_lookup_one_len((name),(parent),(len),__LINE__,__FILE__) +# define RECORD_PATH_LOOKUP(nd) record_path_lookup((nd),__LINE__,__FILE__) +# define RECORD_PATH_RELEASE(nd) record_path_release((nd),__LINE__,__FILE__) +/* This has the effect of reducing the reference count sooner or later, + * if the file is closed. If it isn't then the mount will be busy and + * you can't unmount. + */ +# define DENTRY_OPEN(d,m,f) unionfs_dentry_open((d),(m),(f),__LINE__,__FILE__) +# define GET_PARENT(dentry) unionfs_dget_parent((dentry),__LINE__,__FILE__) +#else /* not FIST_MALLOC_DEBUG */ +# define KZALLOC(a,b) kzalloc((a),(b)) +# define KMALLOC(a,b) kmalloc((a),(b)) +# define KFREE(a) kfree((a)) +# define DPUT(a) dput((a)) +# define DGET(a) dget((a)) +# define IPUT(a) iput((a)) +# define IGRAB(a) igrab((a)) +# define LOOKUP_ONE_LEN(a,b,c) lookup_one_len((a),(b),(c)) +# define RECORD_PATH_LOOKUP(a) +# define RECORD_PATH_RELEASE(a) +# define DENTRY_OPEN(d,m,f) dentry_open((d),(m),(f)) +# define GET_PARENT(d) dget_parent(d) +#endif /* not FIST_MALLOC_DEBUG */ + +/* We can only use 32-bits of offset for rdstate --- blech! */ +#define DIREOF (0xfffff) +#define RDOFFBITS 20 /* This is the number of bits in DIREOF. */ +#define MAXRDCOOKIE (0xfff) +/* Turn an rdstate into an offset. */ +static inline off_t rdstate2offset(struct unionfs_dir_state *buf) +{ + off_t tmp; + tmp = + ((buf->uds_cookie & MAXRDCOOKIE) << RDOFFBITS) | (buf-> + uds_offset & + DIREOF); + return tmp; +} + +#define unionfs_read_lock(sb) down_read(&stopd(sb)->usi_rwsem) +#define unionfs_read_unlock(sb) up_read(&stopd(sb)->usi_rwsem) +#define unionfs_write_lock(sb) down_write(&stopd(sb)->usi_rwsem) +#define unionfs_write_unlock(sb) up_write(&stopd(sb)->usi_rwsem) + +/* The double lock function needs to go after the debugmacros, so that + * dtopd is defined. */ +static inline void double_lock_dentry(struct dentry *d1, struct dentry *d2) +{ + if (d2 < d1) { + struct dentry *tmp = d1; + d1 = d2; + d2 = tmp; + } + lock_dentry(d1); + lock_dentry(d2); +} + +extern int new_dentry_private_data(struct dentry *dentry); +void free_dentry_private_data(struct unionfs_dentry_info *udi); +void update_bstart(struct dentry *dentry); +#define sbt(sb) ((sb)->s_type->name) + +/* + * EXTERNALS: + */ +/* replicates the directory structure upto given dentry in given branch */ +extern struct dentry *create_parents(struct inode *dir, struct dentry *dentry, + int bindex); +struct dentry *create_parents_named(struct inode *dir, struct dentry *dentry, + const char *name, int bindex); + +/* check if two branches overlap */ +extern int is_branch_overlap(struct dentry *dent1, struct dentry *dent2); + +/* partial lookup */ +extern int unionfs_partial_lookup(struct dentry *dentry); + +/* Pass an unionfs dentry and an index and it will try to create a whiteout in branch 'index'. + On error, it will proceed to a branch to the left */ +extern int create_whiteout(struct dentry *dentry, int start); +/* copies a file from dbstart to newbindex branch */ +extern int copyup_file(struct inode *dir, struct file *file, int bstart, + int newbindex, loff_t size); +extern int copyup_named_file(struct inode *dir, struct file *file, + char *name, int bstart, int new_bindex, + loff_t len); + +/* copies a dentry from dbstart to newbindex branch */ +extern int copyup_dentry(struct inode *dir, struct dentry *dentry, int bstart, + int new_bindex, struct file **copyup_file, loff_t len); +extern int copyup_named_dentry(struct inode *dir, struct dentry *dentry, + int bstart, int new_bindex, const char *name, + int namelen, struct file **copyup_file, + loff_t len); + +extern int remove_whiteouts(struct dentry *dentry, struct dentry *hidden_dentry, + int bindex); + +/* Is this directory empty: 0 if it is empty, -ENOTEMPTY if not. */ +extern int check_empty(struct dentry *dentry, + struct unionfs_dir_state **namelist); +/* Delete whiteouts from this directory in branch bindex. */ +extern int delete_whiteouts(struct dentry *dentry, int bindex, + struct unionfs_dir_state *namelist); + +/* Re-lookup a hidden dentry. */ +extern int unionfs_refresh_hidden_dentry(struct dentry *dentry, int bindex); + +extern void unionfs_reinterpose(struct dentry *this_dentry); +extern struct super_block *unionfs_duplicate_super(struct super_block *sb); + +/* Locking functions. */ +extern int unionfs_setlk(struct file *file, int cmd, struct file_lock *fl); +extern int unionfs_getlk(struct file *file, struct file_lock *fl); + +/* Common file operations. */ +extern int unionfs_file_revalidate(struct file *file, int willwrite); +extern int unionfs_open(struct inode *inode, struct file *file); +extern int unionfs_file_release(struct inode *inode, struct file *file); +extern int unionfs_flush(struct file *file, fl_owner_t id); +extern long unionfs_ioctl(struct file *file, unsigned int cmd, + unsigned long arg); + +/* Inode operations */ +extern int unionfs_rename(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry); +int unionfs_unlink(struct inode *dir, struct dentry *dentry); +int unionfs_rmdir(struct inode *dir, struct dentry *dentry); + +int unionfs_d_revalidate(struct dentry *dentry, struct nameidata *nd); + +/* The values for unionfs_interpose's flag. */ +#define INTERPOSE_DEFAULT 0 +#define INTERPOSE_LOOKUP 1 +#define INTERPOSE_REVAL 2 +#define INTERPOSE_REVAL_NEG 3 +#define INTERPOSE_PARTIAL 4 + +extern int unionfs_interpose(struct dentry *this_dentry, struct super_block *sb, + int flag); + +/* Branch management ioctls. */ +int unionfs_ioctl_branchcount(struct file *file, unsigned int cmd, + unsigned long arg); +int unionfs_ioctl_incgen(struct file *file, unsigned int cmd, + unsigned long arg); +int unionfs_ioctl_addbranch(struct inode *inode, unsigned int cmd, + unsigned long arg); +int unionfs_ioctl_delbranch(struct super_block *sb, unsigned long arg); +int unionfs_ioctl_rdwrbranch(struct inode *inode, unsigned int cmd, + unsigned long arg); +int unionfs_ioctl_queryfile(struct file *file, unsigned int cmd, + unsigned long arg); + +/* Verify that a branch is valid. */ +int check_branch(struct nameidata *nd); + +/* Extended attribute functions. */ +extern void *xattr_alloc(size_t size, size_t limit); +extern void xattr_free(void *ptr, size_t size); + +extern ssize_t unionfs_getxattr(struct dentry *dentry, const char *name, + void *value, size_t size); +extern int unionfs_removexattr(struct dentry *dentry, const char *name); +extern ssize_t unionfs_listxattr(struct dentry *dentry, char *list, + size_t size); + +int unionfs_setxattr(struct dentry *dentry, const char *name, const void *value, + size_t size, int flags); + +/* The root directory is unhashed, but isn't deleted. */ +static inline int d_deleted(struct dentry *d) +{ + return d_unhashed(d) && (d != d->d_sb->s_root); +} + +/* returns the sum of the n_link values of all the underlying inodes of the passed inode */ +static inline int get_nlinks(struct inode *inode) +{ + int sum_nlinks = 0; + int dirs = 0; + int bindex; + struct inode *hidden_inode; + + if (!S_ISDIR(inode->i_mode)) + return itohi(inode)->i_nlink; + + for (bindex = ibstart(inode); bindex <= ibend(inode); bindex++) { + hidden_inode = itohi_index(inode, bindex); + if (!hidden_inode || !S_ISDIR(hidden_inode->i_mode)) + continue; + BUG_ON(hidden_inode->i_nlink < 0); + + /* A deleted directory. */ + if (hidden_inode->i_nlink == 0) + continue; + dirs++; + /* A broken directory (e.g., squashfs). */ + if (hidden_inode->i_nlink == 1) + sum_nlinks += 2; + else + sum_nlinks += (hidden_inode->i_nlink - 2); + } + + if (!dirs) + return 0; + return sum_nlinks + 2; +} + +static inline void fist_copy_attr_atime(struct inode *dest, + const struct inode *src) +{ + dest->i_atime = src->i_atime; +} +static inline void fist_copy_attr_times(struct inode *dest, + const struct inode *src) +{ + dest->i_atime = src->i_atime; + dest->i_mtime = src->i_mtime; + dest->i_ctime = src->i_ctime; +} +static inline void fist_copy_attr_timesizes(struct inode *dest, + const struct inode *src) +{ + dest->i_atime = src->i_atime; + dest->i_mtime = src->i_mtime; + dest->i_ctime = src->i_ctime; + dest->i_size = src->i_size; + dest->i_blocks = src->i_blocks; +} +static inline void fist_copy_attr_all(struct inode *dest, + const struct inode *src) +{ + print_entry_location(); + + dest->i_mode = src->i_mode; + /* we do not need to copy if the file is a deleted file */ + if (dest->i_nlink > 0) + dest->i_nlink = get_nlinks(dest); + dest->i_uid = src->i_uid; + dest->i_gid = src->i_gid; + dest->i_rdev = src->i_rdev; + dest->i_atime = src->i_atime; + dest->i_mtime = src->i_mtime; + dest->i_ctime = src->i_ctime; + dest->i_blkbits = src->i_blkbits; + dest->i_size = src->i_size; + dest->i_blocks = src->i_blocks; + dest->i_flags = src->i_flags; + + print_exit_location(); +} + +struct dentry *unionfs_lookup_backend(struct dentry *dentry, struct nameidata *nd, int lookupmode); +int is_stale_inode(struct inode *inode); +void make_stale_inode(struct inode *inode); + +#define IS_SET(sb, check_flag) (check_flag & MOUNT_FLAG(sb)) + +/* unionfs_permission, check if we should bypass error to facilitate copyup */ +#define IS_COPYUP_ERR(err) (err == -EROFS) + +/* unionfs_open, check if we need to copyup the file */ +#define OPEN_WRITE_FLAGS (O_WRONLY | O_RDWR | O_APPEND) +#define IS_WRITE_FLAG(flag) (flag & (OPEN_WRITE_FLAGS)) + +static inline int branchperms(struct super_block *sb, int index) +{ + BUG_ON(index < 0); + + return stopd(sb)->usi_data[index].branchperms; +} +static inline int set_branchperms(struct super_block *sb, int index, int perms) +{ + BUG_ON(index < 0); + + stopd(sb)->usi_data[index].branchperms = perms; + + return perms; +} + +/* Is this file on a read-only branch? */ +static inline int __is_robranch_super(struct super_block *sb, int index, + char *file, const char *function, + int line) +{ + int err = 0; + + print_util_entry_location(); + + if (!(branchperms(sb, index) & MAY_WRITE)) + err = -EROFS; + + print_util_exit_status(err); + return err; +} + +/* Is this file on a read-only branch? */ +static inline int __is_robranch_index(struct dentry *dentry, int index, + char *file, const char *function, + int line) +{ + int err = 0; + int perms; + + print_util_entry_location(); + + BUG_ON(index < 0); + + perms = stopd(dentry->d_sb)->usi_data[index].branchperms; + + if ((!(perms & MAY_WRITE)) + || (IS_RDONLY(dtohd_index(dentry, index)->d_inode))) + err = -EROFS; + + print_util_exit_status(err); + + return err; +} +static inline int __is_robranch(struct dentry *dentry, char *file, + const char *function, int line) +{ + int index; + int err; + + print_util_entry_location(); + + index = dtopd(dentry)->udi_bstart; + BUG_ON(index < 0); + + err = __is_robranch_index(dentry, index, file, function, line); + + print_util_exit_status(err); + + return err; +} + +#define is_robranch(d) __is_robranch(d, __FILE__, __FUNCTION__, __LINE__) +#define is_robranch_super(s, n) __is_robranch_super(s, n, __FILE__, __FUNCTION__, __LINE__) + +/* What do we use for whiteouts. */ +#define WHPFX ".wh." +#define WHLEN 4 +/* If a directory contains this file, then it is opaque. We start with the + * .wh. flag so that it is blocked by loomkup. + */ +#define UNIONFS_DIR_OPAQUE_NAME "__dir_opaque" +#define UNIONFS_DIR_OPAQUE WHPFX UNIONFS_DIR_OPAQUE_NAME + +/* construct whiteout filename */ +static inline char *alloc_whname(const char *name, int len) +{ + char *buf; + + buf = KMALLOC(len + WHLEN + 1, GFP_KERNEL); + if (!buf) + return ERR_PTR(-ENOMEM); + + strcpy(buf, WHPFX); + strlcat(buf, name, len + WHLEN + 1); + + return buf; +} + +/* Definitions for various ways to handle errors. + Each flag's value is its bit position */ + +/* 1 = DELETE_ALL, 0 = check for DELETE_WHITEOUT */ +#ifdef UNIONFS_DELETE_ALL +#define DELETE_ALL 4 +#else +#define DELETE_ALL 0 +#endif + +#define VALID_MOUNT_FLAGS (DELETE_ALL) + +/* + * MACROS: + */ + +#ifndef SEEK_SET +#define SEEK_SET 0 +#endif /* not SEEK_SET */ + +#ifndef SEEK_CUR +#define SEEK_CUR 1 +#endif /* not SEEK_CUR */ + +#ifndef SEEK_END +#define SEEK_END 2 +#endif /* not SEEK_END */ + +#ifndef DEFAULT_POLLMASK +#define DEFAULT_POLLMASK (POLLIN | POLLOUT | POLLRDNORM | POLLWRNORM) +#endif + +/* + * EXTERNALS: + */ + +/* JS: These two functions are here because it is kind of daft to copy and paste the + * contents of the two functions to 32+ places in unionfs + */ +static inline struct dentry *lock_parent(struct dentry *dentry) +{ + struct dentry *dir = DGET(dentry->d_parent); + + mutex_lock(&dir->d_inode->i_mutex); + return dir; +} + +static inline void unlock_dir(struct dentry *dir) +{ + mutex_unlock(&dir->d_inode->i_mutex); + DPUT(dir); +} + +extern int make_dir_opaque(struct dentry *dir, int bindex); + +#endif /* __KERNEL__ */ + +/* + * DEFINITIONS FOR USER AND KERNEL CODE: + * (Note: ioctl numbers 1--9 are reserved for fistgen, the rest + * are auto-generated automatically based on the user's .fist file.) + */ +# define FIST_IOCTL_GET_DEBUG_VALUE _IOR(0x15, 1, int) +# define FIST_IOCTL_SET_DEBUG_VALUE _IOW(0x15, 2, int) +# define UNIONFS_IOCTL_BRANCH_COUNT _IOR(0x15, 10, int) +# define UNIONFS_IOCTL_INCGEN _IOR(0x15, 11, int) +# define UNIONFS_IOCTL_ADDBRANCH _IOW(0x15, 12, int) +# define UNIONFS_IOCTL_DELBRANCH _IOW(0x15, 13, int) +# define UNIONFS_IOCTL_RDWRBRANCH _IOW(0x15, 14, int) +# define UNIONFS_IOCTL_QUERYFILE _IOR(0x15, 15, int) + +/* We don't support normal remount, but unionctl uses it. */ +# define UNIONFS_REMOUNT_MAGIC 0x4a5a4380 + +/* should be at least LAST_USED_UNIONFS_PERMISSION<<1 */ +#define MAY_NFSRO 16 + +struct unionfs_addbranch_args { + unsigned int ab_branch; + char *ab_path; + unsigned int ab_perms; +}; + +struct unionfs_rdwrbranch_args { + unsigned int rwb_branch; + unsigned int rwb_perms; +}; + +#endif /* not __UNIONFS_H_ */ +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/unionfs_debug.h @@ -0,0 +1,191 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef Sipek + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: unionfs_debug.h,v 1.3 2006/06/01 21:25:18 jsipek Exp $ + */ + +#ifndef __UNIONFS_H_ +#error This file should only be included from unionfs.h! +#endif + +#ifdef UNIONFS_DEBUG +#define DEFAULT_DEBUG_MASK 0 +#else +#define DEFAULT_DEBUG_MASK (~0) +#endif + +/* debug print levels */ +#define PRINT_NONE 0x0000 +#define PRINT_MAIN_ENTRY 0x0001 +#define PRINT_MAIN_EXIT 0x0002 +#define PRINT_UTILITY_ENTRY 0x0004 +#define PRINT_UTILITY_EXIT 0x0008 +#define PRINT_MISC_ENTRY 0x0010 +#define PRINT_MISC_EXIT 0x0020 +#define PRINT_DATA_DENTRY 0x0040 +#define PRINT_DATA_FILE 0x0080 +#define PRINT_DATA_INODE 0x0100 +#define PRINT_DATA_SB 0x0200 +#define PRINT_DEBUG 0x0400 +#define __PRINT_DEBUG_XATTR 0x0800 +#define PRINT_DEBUG_XATTR (PRINT_DEBUG | __PRINT_DEBUG_XATTR) +#define __PRINT_DEBUG_WHITEOUT 0x1000 +#define PRINT_DEBUG_WHITEOUT (PRINT_DEBUG | __PRINT_DEBUG_WHITEOUT) + +#define PRINT_MAX (0x2000 - 1) +#define PRINT_ALL (~PRINT_NONE) + +extern unsigned int get_debug_mask(void); +extern int set_debug_mask(int val); + +/* print inode */ +extern void unionfs_print_inode(const unsigned int req, const char *prefix, const struct inode *inode); + +/* check inode */ +extern void unionfs_checkinode(const unsigned int req, const struct inode *inode, const char *msg); + +/* prunt file */ +extern void unionfs_print_file(const unsigned int req, const char *prefix, const struct file *file); + +/* print dentry */ +extern void unionfs_print_dentry(const unsigned int req, const char *prefix, const struct dentry *dentry); + +extern void unionfs_print_dentry_nocheck(const unsigned int req, const char *prefix, const struct dentry *dentry); + +/* print superblock */ +extern void unionfs_print_sb(const unsigned int req, const char *prefix, const struct super_block *sb); + +/* print message */ +extern int unionfs_print(const unsigned int req, const char *fmt, ...); + +/* forced print-debugging functions */ +#define force_print_dentry(prefix, ptr) \ + unionfs_print_dentry(PRINT_ALL, (prefix), (ptr)) +#define force_print_dentry_nocheck(prefix, ptr) \ + unionfs_print_dentry_nocheck(PRINT_ALL, (prefix), (ptr)) +#define force_print_file(prefix, ptr) \ + unionfs_print_file(PRINT_ALL, (prefix), (ptr)) +#define force_print_inode(prefix, ptr) \ + unionfs_print_inode(PRINT_ALL, (prefix), (ptr)) +#define force_print_sb(prefix, ptr) \ + unionfs_print_sb(PRINT_ALL, (prefix), (ptr)) + +#ifdef UNIONFS_DEBUG +/* + * Full-fledged debugging enabled + */ + +#define print_dentry(prefix, ptr) \ + unionfs_print_dentry(PRINT_DATA_DENTRY, (prefix), (ptr)) +#define print_dentry_nocheck(prefix, ptr) \ + unionfs_print_dentry_nocheck(PRINT_DATA_DENTRY, (prefix), (ptr)) +#define print_file(prefix, ptr) \ + unionfs_print_file(PRINT_DATA_FILE, (prefix), (ptr)) +#define print_inode(prefix, ptr) \ + unionfs_print_inode(PRINT_DATA_INODE, (prefix), (ptr)) +#define print_sb(prefix, ptr) \ + unionfs_print_sb(PRINT_DATA_SB, (prefix), (ptr)) +#define dprint(req, fmt, args...) \ + unionfs_print(req, fmt, ## args) + +#define checkinode(ptr, msg) \ + unionfs_checkinode(PRINT_DEBUG, (ptr), (msg)) + +#define __print_entryexit(req, ee, fmt, args...) \ + unionfs_print((req), \ + ee " %s %s:%d" fmt "\n", \ + __FUNCTION__, \ + __FILE__, \ + __LINE__, \ + ##args) + +#define print_entry(fmt, args...) \ + __print_entryexit(PRINT_MAIN_ENTRY, \ + "IN: ", " " fmt, ##args) + +#define print_entry_location() \ + __print_entryexit(PRINT_MAIN_ENTRY, \ + "IN: ", "") + +#define print_exit_location() \ + __print_entryexit(PRINT_MAIN_EXIT, \ + "OUT:", "") + +#define print_exit_status(status) \ + __print_entryexit(PRINT_MAIN_EXIT, \ + "OUT:", ", STATUS: %d", status) + +static inline void __print_exit_pointer(unsigned int req, void *status) +{ + if (IS_ERR(status)) + __print_entryexit(req, "OUT:", ", STATUS: %ld", + PTR_ERR(status)); + else + __print_entryexit(req, "OUT:", ", STATUS: 0x%p", + status); +} +#define print_exit_pointer(status) \ + __print_exit_pointer(PRINT_MAIN_EXIT, status) + +#define print_util_entry(fmt, args...) \ + __print_entryexit(PRINT_UTILITY_ENTRY, \ + "IN: ", " " fmt, ##args) + +#define print_util_entry_location() \ + __print_entryexit(PRINT_UTILITY_ENTRY, \ + "IN: ", "") + +#define print_util_exit_location() \ + __print_entryexit(PRINT_UTILITY_EXIT, \ + "OUT:", "") + +#define print_util_exit_status(status) \ + __print_entryexit(PRINT_UTILITY_EXIT, \ + "OUT:", ", STATUS: %d", status) + +#define print_util_exit_pointer(status) \ + __print_exit_pointer(PRINT_UTILITY_EXIT, status) + +#else /* UNIONFS_DEBUG */ +/* + * Full-fledged debugging disabled + */ + +#define print_dentry(prefix, ptr) +#define print_dentry_nocheck(prefix, ptr) +#define print_file(prefix, ptr) +#define print_inode(prefix, ptr) +#define print_sb(prefix, ptr) +#define dprint(req, fmt, args...) + +#define checkinode(ptr, msg) + +#define print_entry(args...) +#define print_entry_location() +#define print_exit_location() +#define print_exit_status(status) +#define print_exit_pointer(status) +#define print_util_entry(args...) +#define print_util_entry_location() +#define print_util_exit_location() +#define print_util_exit_status(status) +#define print_util_exit_pointer(status) + +#endif /* ! UNIONFS_DEBUG */ + + --- /dev/null +++ b/fs/unionfs/unionfs_imap.h @@ -0,0 +1,92 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef Sipek + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: unionfs_imap.h,v 1.1 2006/05/30 21:38:45 jsipek Exp $ + */ + +#ifndef __UNIONFS_H_ +#error This file should only be included from unionfs.h! +#endif + +#ifdef UNIONFS_IMAP + +/*UUID typedef needed later*/ +typedef uint8_t uuid_t[16]; + +/* +* Defines,structs,and functions for persistent used by kernel and user +*/ +#define MAX_MAPS 256 +#define UUID_LEN 16 +#define FORWARDMAP_MAGIC 0x4b1cb38f +#define REVERSEMAP_MAGIC 0Xfcafad71 +#define FORWARDMAP_VERSION 0x02 +#define REVERSEMAP_VERSION 0x01 +#define FIRST_VALID_INODE 3 +struct fmaphdr { + uint32_t magic; + uint32_t version; + uint8_t usedbranches; + uint8_t uuid[UUID_LEN]; +}; + +struct rmaphdr { + uint32_t magic; + uint32_t version; + uint8_t fwduuid[UUID_LEN]; + uint8_t revuuid[UUID_LEN]; + fsid_t fsid; +}; +struct bmapent { + fsid_t fsid; + uint8_t uuid[UUID_LEN]; +}; +struct fmapent { + uint8_t fsnum; + uint64_t inode; +}; + +/* Persistant Inode functions */ +extern int read_uin(struct super_block *sb, uint8_t branchnum, + ino_t inode_number, int flag, ino_t * uino); +extern int write_uin(struct super_block *sb, ino_t ino, int bindex, + ino_t hidden_ino); +extern int get_lin(struct super_block *sb, ino_t inode_number, + struct fmapent *entry); +extern int parse_imap_option(struct super_block *sb, + struct unionfs_dentry_info *hidden_root_info, + char *options); +extern void cleanup_imap_data(struct super_block *sb); + +#endif /*#ifdef UNIONFS_IMAP */ + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/unionfs_macros.h @@ -0,0 +1,206 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef Sipek + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: unionfs_macros.h,v 1.13 2006/06/01 03:11:03 jsipek Exp $ + */ + +#ifndef __UNIONFS_H_ +#error This file should only be included from unionfs.h! +#endif + +/* Inode to private data */ +static inline struct unionfs_inode_info *itopd(const struct inode *inode) +{ + return + &(container_of(inode, struct unionfs_inode_container, vfs_inode)-> + info); +} + +#define itohi_ptr(ino) (itopd(ino)->uii_inode) +#define ibstart(ino) (itopd(ino)->b_start) +#define ibend(ino) (itopd(ino)->b_end) + +/* Superblock to private data */ +#define stopd(super) ((struct unionfs_sb_info *)(super)->s_fs_info) +#define stopd_lhs(super) ((super)->s_fs_info) +#define sbstart(sb) 0 +#define sbend(sb) stopd(sb)->b_end +#define sbmax(sb) (stopd(sb)->b_end + 1) + +/* File to private Data */ +#define ftopd(file) ((struct unionfs_file_info *)((file)->private_data)) +#define ftopd_lhs(file) ((file)->private_data) +#define ftohf_ptr(file) (ftopd(file)->ufi_file) +#define fbstart(file) (ftopd(file)->b_start) +#define fbend(file) (ftopd(file)->b_end) + +/* File to hidden file. */ +static inline struct file *ftohf(struct file *f) +{ + return ftopd(f)->ufi_file[fbstart(f)]; +} + +static inline struct file *ftohf_index(const struct file *f, int index) +{ + return ftopd(f)->ufi_file[index]; +} + +static inline void set_ftohf_index(struct file *f, int index, struct file *val) +{ + ftopd(f)->ufi_file[index] = val; +} + +static inline void set_ftohf(struct file *f, struct file *val) +{ + ftopd(f)->ufi_file[fbstart(f)] = val; +} + +/* Inode to hidden inode. */ +static inline struct inode *itohi(const struct inode *i) +{ + return itopd(i)->uii_inode[ibstart(i)]; +} + +static inline struct inode *itohi_index(const struct inode *i, int index) +{ + return itopd(i)->uii_inode[index]; +} + +static inline void set_itohi_index(struct inode *i, int index, + struct inode *val) +{ + itopd(i)->uii_inode[index] = val; +} + +static inline void set_itohi(struct inode *i, struct inode *val) +{ + itopd(i)->uii_inode[ibstart(i)] = val; +} + +/* Superblock to hidden superblock. */ +static inline struct super_block *stohs(const struct super_block *o) +{ + return stopd(o)->usi_data[sbstart(o)].sb; +} + +static inline struct super_block *stohs_index(const struct super_block *o, int index) +{ + return stopd(o)->usi_data[index].sb; +} + +static inline void set_stohs_index(struct super_block *o, int index, + struct super_block *val) +{ + stopd(o)->usi_data[index].sb = val; +} + +static inline void set_stohs(struct super_block *o, struct super_block *val) +{ + stopd(o)->usi_data[sbstart(o)].sb = val; +} + +/* Super to hidden mount. */ +static inline struct vfsmount *stohiddenmnt_index(struct super_block *o, + int index) +{ + return stopd(o)->usi_data[index].hidden_mnt; +} + +static inline void set_stohiddenmnt_index(struct super_block *o, int index, + struct vfsmount *val) +{ + stopd(o)->usi_data[index].hidden_mnt = val; +} + +/* Branch count macros. */ +static inline int branch_count(struct super_block *o, int index) +{ + return atomic_read(&stopd(o)->usi_data[index].sbcount); +} + +static inline void set_branch_count(struct super_block *o, int index, int val) +{ + atomic_set(&stopd(o)->usi_data[index].sbcount, val); +} + +static inline void branchget(struct super_block *o, int index) +{ + atomic_inc(&stopd(o)->usi_data[index].sbcount); +} + +static inline void branchput(struct super_block *o, int index) +{ + atomic_dec(&stopd(o)->usi_data[index].sbcount); +} + +/* Dentry macros */ +static inline struct unionfs_dentry_info *dtopd(const struct dentry *dent) +{ + return (struct unionfs_dentry_info *)dent->d_fsdata; +} + +#define dtopd_lhs(dent) ((dent)->d_fsdata) +#define dtopd_nocheck(dent) dtopd(dent) +#define dbstart(dent) (dtopd(dent)->udi_bstart) +#define set_dbstart(dent, val) do { dtopd(dent)->udi_bstart = val; } while(0) +#define dbend(dent) (dtopd(dent)->udi_bend) +#define set_dbend(dent, val) do { dtopd(dent)->udi_bend = val; } while(0) +#define dbopaque(dent) (dtopd(dent)->udi_bopaque) +#define set_dbopaque(dent, val) do { dtopd(dent)->udi_bopaque = val; } while (0) + +static inline void set_dtohd_index(struct dentry *dent, int index, + struct dentry *val) +{ + dtopd(dent)->udi_dentry[index] = val; +} + +static inline struct dentry *dtohd_index(const struct dentry *dent, int index) +{ + return dtopd(dent)->udi_dentry[index]; +} + +static inline struct dentry *dtohd(const struct dentry *dent) +{ + return dtopd(dent)->udi_dentry[dbstart(dent)]; +} + +#define set_dtohd_index_nocheck(dent, index, val) set_dtohd_index(dent, index, val) +#define dtohd_index_nocheck(dent, index) dtohd_index(dent, index) + +#define dtohd_ptr(dent) (dtopd_nocheck(dent)->udi_dentry) + +/* Macros for locking a dentry. */ +#define lock_dentry(d) down(&dtopd(d)->udi_sem) +#define unlock_dentry(d) up(&dtopd(d)->udi_sem) +#define verify_locked(d) + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/unlink.c @@ -0,0 +1,380 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef 'Jeff' Sipek + * Copyright (c) 2005-2006 Junjiro Okajima + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: unlink.c,v 1.44 2006/08/05 01:28:46 jro Exp $ + */ + +#include "unionfs.h" + +#ifdef UNIONFS_DELETE_ALL +static int unionfs_unlink_all(struct inode *dir, struct dentry *dentry) +{ + struct dentry *hidden_dentry; + struct dentry *hidden_dir_dentry; + int bstart, bend, bindex; + int err = 0; + int global_err = 0; + + print_entry_location(); + + if ((err = unionfs_partial_lookup(dentry))) + goto out; + + bstart = dbstart(dentry); + bend = dbend(dentry); + + for (bindex = bend; bindex >= bstart; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + + hidden_dir_dentry = lock_parent(hidden_dentry); + + /* avoid destroying the hidden inode if the file is in use */ + DGET(hidden_dentry); + if (!(err = is_robranch_super(dentry->d_sb, bindex))) + err = vfs_unlink(hidden_dir_dentry->d_inode, + hidden_dentry, NULL); + DPUT(hidden_dentry); + fist_copy_attr_times(dir, hidden_dir_dentry->d_inode); + unlock_dir(hidden_dir_dentry); + + if (err) { + /* passup the last error we got */ + if (!IS_COPYUP_ERR(err)) + goto out; + global_err = err; + } + } + + /* check if encountered error in the above loop */ + if (global_err) { + /* If we failed in the leftmost branch, then err will be set + * and we should move one over to create the whiteout. + * Otherwise, we should try in the leftmost branch. */ + if (err) { + if (dbstart(dentry) == 0) { + goto out; + } + err = create_whiteout(dentry, dbstart(dentry) - 1); + } else { + err = create_whiteout(dentry, dbstart(dentry)); + } + } else if (dbopaque(dentry) != -1) { + /* There is a hidden lower-priority file with the same name. */ + err = create_whiteout(dentry, dbopaque(dentry)); + } + out: + /* propagate number of hard-links */ + if (dentry->d_inode->i_nlink != 0) { + dentry->d_inode->i_nlink = get_nlinks(dentry->d_inode); + if (!err && global_err) + dentry->d_inode->i_nlink--; + } + /* We don't want to leave negative leftover dentries for revalidate. */ + if (!err && (global_err || dbopaque(dentry) != -1)) + update_bstart(dentry); + + print_exit_status(err); + return err; +} +#endif +static int unionfs_unlink_whiteout(struct inode *dir, struct dentry *dentry) +{ + struct dentry *hidden_dentry; + struct dentry *hidden_dir_dentry; + int bindex; + int err = 0; + + print_entry_location(); + + if ((err = unionfs_partial_lookup(dentry))) + goto out; + + bindex = dbstart(dentry); + + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + goto out; + + hidden_dir_dentry = lock_parent(hidden_dentry); + + /* avoid destroying the hidden inode if the file is in use */ + DGET(hidden_dentry); + if (!(err = is_robranch_super(dentry->d_sb, bindex))) + err = vfs_unlink(hidden_dir_dentry->d_inode, hidden_dentry, + NULL); + DPUT(hidden_dentry); + fist_copy_attr_times(dir, hidden_dir_dentry->d_inode); + unlock_dir(hidden_dir_dentry); + + if (err && !IS_COPYUP_ERR(err)) + goto out; + + if (err) { + if (dbstart(dentry) == 0) + goto out; + + err = create_whiteout(dentry, dbstart(dentry) - 1); + } else if (dbopaque(dentry) != -1) { + /* There is a hidden lower-priority file with the same name. */ + err = create_whiteout(dentry, dbopaque(dentry)); + } else { + err = create_whiteout(dentry, dbstart(dentry)); + } + + out: + if (!err) + dentry->d_inode->i_nlink--; + + /* We don't want to leave negative leftover dentries for revalidate. */ + if (!err && (dbopaque(dentry) != -1)) + update_bstart(dentry); + + print_exit_status(err); + return err; + +} + +int unionfs_unlink(struct inode *dir, struct dentry *dentry) +{ + int err = 0; + + print_entry_location(); + lock_dentry(dentry); + print_dentry("IN unionfs_unlink", dentry); + +#ifdef UNIONFS_DELETE_ALL + if (IS_SET(dir->i_sb, DELETE_ALL)) + err = unionfs_unlink_all(dir, dentry); + else +#endif + err = unionfs_unlink_whiteout(dir, dentry); + /* call d_drop so the system "forgets" about us */ + if (!err) + d_drop(dentry); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +static int unionfs_rmdir_first(struct inode *dir, struct dentry *dentry, + struct unionfs_dir_state *namelist) +{ + int err; + struct dentry *hidden_dentry; + struct dentry *hidden_dir_dentry = NULL; + + print_entry_location(); + print_dentry("IN unionfs_rmdir_first: ", dentry); + + /* Here we need to remove whiteout entries. */ + err = delete_whiteouts(dentry, dbstart(dentry), namelist); + if (err) { + goto out; + } + + hidden_dentry = dtohd(dentry); + + hidden_dir_dentry = lock_parent(hidden_dentry); + + /* avoid destroying the hidden inode if the file is in use */ + DGET(hidden_dentry); + if (!(err = is_robranch(dentry))) { + err = vfs_rmdir(hidden_dir_dentry->d_inode, hidden_dentry, + NULL); + } + DPUT(hidden_dentry); + + fist_copy_attr_times(dir, hidden_dir_dentry->d_inode); + /* propagate number of hard-links */ + dentry->d_inode->i_nlink = get_nlinks(dentry->d_inode); + + out: + if (hidden_dir_dentry) { + unlock_dir(hidden_dir_dentry); + } + print_dentry("OUT unionfs_rmdir_first: ", dentry); + print_exit_status(err); + return err; +} + +#ifdef UNIONFS_DELETE_ALL +static int unionfs_rmdir_all(struct inode *dir, struct dentry *dentry, + struct unionfs_dir_state *namelist) +{ + struct dentry *hidden_dentry; + struct dentry *hidden_dir_dentry; + int bstart, bend, bindex; + int err = 0; + int global_err = 0; + + print_entry_location(); + print_dentry("IN unionfs_rmdir_all: ", dentry); + + bstart = dbstart(dentry); + bend = dbend(dentry); + + for (bindex = bend; bindex >= bstart; bindex--) { + hidden_dentry = dtohd_index(dentry, bindex); + if (!hidden_dentry) + continue; + + hidden_dir_dentry = lock_parent(hidden_dentry); + if (S_ISDIR(hidden_dentry->d_inode->i_mode)) { + err = delete_whiteouts(dentry, bindex, namelist); + if (!err + && !(err = + is_robranch_super(dentry->d_sb, bindex))) { + err = + vfs_rmdir(hidden_dir_dentry->d_inode, + hidden_dentry, NULL); + } + } else { + err = -EISDIR; + } + + fist_copy_attr_times(dir, hidden_dir_dentry->d_inode); + unlock_dir(hidden_dir_dentry); + if (err) { + int local_err = + unionfs_refresh_hidden_dentry(dentry, bindex); + if (local_err) { + err = local_err; + goto out; + } + + if (!IS_COPYUP_ERR(err) && err != -ENOTEMPTY + && err != -EISDIR) + goto out; + + global_err = err; + } + } + + /* check if encountered error in the above loop */ + if (global_err) { + /* If we failed in the leftmost branch, then err will be set and we should + * move one over to create the whiteout. Otherwise, we should try in the + * leftmost branch. + */ + if (err) { + if (dbstart(dentry) == 0) { + goto out; + } + err = create_whiteout(dentry, dbstart(dentry) - 1); + } else { + err = create_whiteout(dentry, dbstart(dentry)); + } + } else { + err = create_whiteout(dentry, dbstart(dentry)); + } + + out: + /* propagate number of hard-links */ + dentry->d_inode->i_nlink = get_nlinks(dentry->d_inode); + + print_dentry("OUT unionfs_rmdir_all: ", dentry); + print_exit_status(err); + return err; +} +#endif +int unionfs_rmdir(struct inode *dir, struct dentry *dentry) +{ + int err = 0; + struct unionfs_dir_state *namelist = NULL; + + print_entry_location(); + lock_dentry(dentry); + print_dentry("IN unionfs_rmdir: ", dentry); + + /* check if this unionfs directory is empty or not */ + err = check_empty(dentry, &namelist); + if (err) { +#if 0 + /* vfs_rmdir(our caller) unhashed the dentry. This will recover + * the Unionfs inode number for the directory itself, but the + * children are already lost. It seems that tmpfs manages its + * way around this by upping the refcount on everything. + * + * Even if we do this, we still lose the inode numbers of the + * children. The best way to fix this is to fix the VFS (or + * use persistent inode maps). */ + if (d_unhashed(dentry)) + d_rehash(dentry); +#endif + goto out; + } +#ifdef UNIONFS_DELETE_ALL + if (IS_SET(dir->i_sb, DELETE_ALL)) { + /* delete all. */ + err = unionfs_rmdir_all(dir, dentry, namelist); + } else { /* Delete the first directory. */ +#endif + err = unionfs_rmdir_first(dir, dentry, namelist); + /* create whiteout */ + if (!err) { + err = create_whiteout(dentry, dbstart(dentry)); + } else { + int new_err; + + if (dbstart(dentry) == 0) + goto out; + + /* exit if the error returned was NOT -EROFS */ + if (!IS_COPYUP_ERR(err)) + goto out; + + new_err = create_whiteout(dentry, dbstart(dentry) - 1); + if (new_err != -EEXIST) + err = new_err; + } + +#ifdef UNIONFS_DELETE_ALL + } +#endif + out: + /* call d_drop so the system "forgets" about us */ + if (!err) + d_drop(dentry); + + if (namelist) + free_rdstate(namelist); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */ --- /dev/null +++ b/fs/unionfs/xattr.c @@ -0,0 +1,165 @@ +/* + * Copyright (c) 2003-2006 Erez Zadok + * Copyright (c) 2003-2006 Charles P. Wright + * Copyright (c) 2005-2006 Josef Sipek + * Copyright (c) 2005 Arun M. Krishnakumar + * Copyright (c) 2005-2006 David P. Quigley + * Copyright (c) 2003-2004 Mohammad Nayyer Zubair + * Copyright (c) 2003 Puja Gupta + * Copyright (c) 2003 Harikesavan Krishnan + * Copyright (c) 2003-2006 Stony Brook University + * Copyright (c) 2003-2006 The Research Foundation of State University of New York + * + * For specific licensing information, see the COPYING file distributed with + * this package. + * + * This Copyright notice must be kept intact and distributed with all sources. + */ +/* + * $Id: xattr.c,v 1.32 2006/06/01 03:11:03 jsipek Exp $ + */ + +#include "unionfs.h" + +/* This is lifted from fs/xattr.c */ +void *xattr_alloc(size_t size, size_t limit) +{ + void *ptr; + + if (size > limit) + return ERR_PTR(-E2BIG); + + if (!size) /* size request, no buffer is needed */ + return NULL; + else if (size <= PAGE_SIZE) + ptr = KMALLOC((unsigned long)size, GFP_KERNEL); + else + ptr = vmalloc((unsigned long)size); + if (!ptr) + return ERR_PTR(-ENOMEM); + return ptr; +} + +void xattr_free(void *ptr, size_t size) +{ + if (!size) /* size request, no buffer was needed */ + return; + else if (size <= PAGE_SIZE) + KFREE(ptr); + else + vfree(ptr); +} + +/* BKL held by caller. + * dentry->d_inode->i_mutex locked + * ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t); + */ +ssize_t unionfs_getxattr(struct dentry * dentry, const char *name, void *value, + size_t size) +{ + struct dentry *hidden_dentry = NULL; + int err = -EOPNOTSUPP; + + print_entry_location(); + + dprint(PRINT_DEBUG_XATTR, "getxattr: name=\"%s\", value %lu bytes\n", + name, size); + + lock_dentry(dentry); + + hidden_dentry = dtohd(dentry); + + err = vfs_getxattr(hidden_dentry, NULL, (char*)name, value, size, + NULL); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +/* BKL held by caller. + * dentry->d_inode->i_mutex locked + */ +int +unionfs_setxattr(struct dentry *dentry, const char *name, const void *value, + size_t size, int flags) +{ + struct dentry *hidden_dentry = NULL; + int err = -EOPNOTSUPP; + + print_entry_location(); + + lock_dentry(dentry); + hidden_dentry = dtohd(dentry); + + dprint(PRINT_DEBUG_XATTR, "setxattr: name=\"%s\", value %lu bytes," + "flags=%x\n", name, (unsigned long)size, flags); + + err = + vfs_setxattr(hidden_dentry, NULL, (char *)name, (char *)value, + size, flags, NULL); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +/* BKL held by caller. + * dentry->d_inode->i_mutex locked + */ +int unionfs_removexattr(struct dentry *dentry, const char *name) +{ + struct dentry *hidden_dentry = NULL; + int err = -EOPNOTSUPP; + + print_entry_location(); + + lock_dentry(dentry); + hidden_dentry = dtohd(dentry); + + dprint(PRINT_DEBUG_XATTR, "removexattr: name=\"%s\"\n", name); + + err = vfs_removexattr(hidden_dentry, NULL, (char*)name, NULL); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +/* BKL held by caller. + * dentry->d_inode->i_mutex locked + */ +ssize_t unionfs_listxattr(struct dentry * dentry, char *list, size_t size) +{ + struct dentry *hidden_dentry = NULL; + int err = -EOPNOTSUPP; + char *encoded_list = NULL; + + print_entry_location(); + lock_dentry(dentry); + + hidden_dentry = dtohd(dentry); + + encoded_list = list; + err = vfs_listxattr(hidden_dentry, NULL, encoded_list, size, NULL); + + unlock_dentry(dentry); + print_exit_status(err); + return err; +} + +/* + * + * vim:shiftwidth=8 + * vim:tabstop=8 + * + * For Emacs: + * Local variables: + * c-basic-offset: 8 + * c-comment-only-line-offset: 0 + * c-offsets-alist: ((statement-block-intro . +) (knr-argdecl-intro . 0) + * (substatement-open . 0) (label . 0) (statement-cont . +)) + * indent-tabs-mode: t + * tab-width: 8 + * End: + */