[AUFS PATCH v2.6.26-rc2-mm1 06/39] aufs branch directory/filesystem

From: hooanon05
Date: Tue May 20 2008 - 23:49:36 EST


From: Junjiro Okajima <hooanon05@xxxxxxxxxxx>

branch management

Signed-off-by: Junjiro Okajima <hooanon05@xxxxxxxxxxx>
---
fs/aufs/branch.c | 912 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
fs/aufs/branch.h | 352 +++++++++++++++++++++
2 files changed, 1264 insertions(+), 0 deletions(-)
create mode 100644 fs/aufs/branch.c
create mode 100644 fs/aufs/branch.h

diff --git a/fs/aufs/branch.c b/fs/aufs/branch.c
new file mode 100644
index 0000000..88b6258
--- /dev/null
+++ b/fs/aufs/branch.c
@@ -0,0 +1,912 @@
+/*
+ * Copyright (C) 2005-2008 Junjiro Okajima
+ *
+ * This program, aufs is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/*
+ * branch management
+ */
+
+#include <linux/iso_fs.h>
+#include <linux/loop.h>
+#include <linux/romfs_fs.h>
+#include <linux/smp_lock.h>
+#include "aufs.h"
+
+static void free_branch(struct au_branch *br)
+{
+ AuTraceEnter();
+
+ if (br->br_xino)
+ fput(br->br_xino);
+ dput(br->br_wh);
+ dput(br->br_plink);
+ if (!au_test_nfs(br->br_mnt->mnt_sb))
+ mntput(br->br_mnt);
+ else {
+ lockdep_off();
+ mntput(br->br_mnt);
+ lockdep_on();
+ }
+ AuDebugOn(au_br_count(br) || atomic_read(&br->br_wh_running));
+ kfree(br);
+}
+
+/*
+ * frees all branches
+ */
+void au_br_free(struct au_sbinfo *sbinfo)
+{
+ aufs_bindex_t bmax;
+ struct au_branch **br;
+
+ AuTraceEnter();
+ bmax = sbinfo->si_bend + 1;
+ br = sbinfo->si_branch;
+ while (bmax--)
+ free_branch(*br++);
+}
+
+/*
+ * find the index of a branch which is specified by @br_id.
+ */
+int au_br_index(struct super_block *sb, aufs_bindex_t br_id)
+{
+ aufs_bindex_t bindex, bend;
+
+ AuTraceEnter();
+
+ bend = au_sbend(sb);
+ for (bindex = 0; bindex <= bend; bindex++)
+ if (au_sbr_id(sb, bindex) == br_id)
+ return bindex;
+ return -1;
+}
+
+/*
+ * test if the @h_sb is real-readonly.
+ */
+int au_test_def_rr(struct super_block *h_sb)
+{
+ switch (h_sb->s_magic) {
+#if defined(CONFIG_ISO9660_FS) || defined(CONFIG_ISO9660_FS_MODULE)
+ case ISOFS_SUPER_MAGIC:
+ return 1;
+#endif
+
+#if defined(CONFIG_CRAMFS) || defined(CONFIG_CRAMFS_MODULE)
+ case CRAMFS_MAGIC:
+ return 1;
+#endif
+
+#if defined(CONFIG_ROMFS_FS) || defined(CONFIG_ROMFS_FS_MODULE)
+ case ROMFS_MAGIC:
+ return 1;
+#endif
+
+ default:
+ return 0;
+ }
+}
+
+/* ---------------------------------------------------------------------- */
+
+/*
+ * test if two hidden_dentries have overlapping branches.
+ */
+static int do_test_overlap(struct super_block *sb, struct dentry *h_d1,
+ struct dentry *h_d2)
+{
+ int err;
+
+ LKTRTrace("%.*s, %.*s\n", AuDLNPair(h_d1), AuDLNPair(h_d2));
+
+ err = au_test_subdir(h_d1, h_d2);
+ AuTraceErr(err);
+ return err;
+}
+
+static int test_overlap_loopback(struct super_block *sb, struct dentry *h_d1,
+ struct dentry *h_d2)
+{
+#if defined(CONFIG_BLK_DEV_LOOP) || defined(CONFIG_BLK_DEV_LOOP_MODULE)
+ struct inode *h_inode;
+ struct loop_device *l;
+
+ h_inode = h_d1->d_inode;
+ if (MAJOR(h_inode->i_sb->s_dev) != LOOP_MAJOR)
+ return 0;
+
+ l = h_inode->i_sb->s_bdev->bd_disk->private_data;
+ h_d1 = l->lo_backing_file->f_dentry;
+ if (unlikely(h_d1->d_sb == sb))
+ return 1;
+ return do_test_overlap(sb, h_d1, h_d2);
+#else
+ return 0;
+#endif
+}
+
+static int test_overlap(struct super_block *sb, struct dentry *h_d1,
+ struct dentry *h_d2)
+{
+ LKTRTrace("d1 %.*s, d2 %.*s\n", AuDLNPair(h_d1), AuDLNPair(h_d2));
+
+ if (unlikely(h_d1 == h_d2))
+ return 1;
+ return do_test_overlap(sb, h_d1, h_d2)
+ || do_test_overlap(sb, h_d2, h_d1)
+ || test_overlap_loopback(sb, h_d1, h_d2)
+ || test_overlap_loopback(sb, h_d2, h_d1);
+}
+
+/* ---------------------------------------------------------------------- */
+
+static int init_br_wh(struct super_block *sb, aufs_bindex_t bindex,
+ struct au_branch *br, int new_perm,
+ struct dentry *h_root, struct vfsmount *h_mnt)
+{
+ int err, old_perm;
+ struct inode *dir, *h_dir;
+ const int new = (bindex < 0);
+
+ LKTRTrace("b%d, new_perm %d\n", bindex, new_perm);
+
+ dir = sb->s_root->d_inode;
+ h_dir = h_root->d_inode;
+ if (new)
+ mutex_lock_nested(&h_dir->i_mutex, AuLsc_I_PARENT);
+ else
+ au_hdir_lock(h_dir, dir, bindex);
+
+ br_wh_write_lock(br);
+ old_perm = br->br_perm;
+ br->br_perm = new_perm;
+ err = au_wh_init(h_root, br, au_do_nfsmnt(h_mnt), sb);
+ br->br_perm = old_perm;
+ br_wh_write_unlock(br);
+
+ if (new)
+ mutex_unlock(&h_dir->i_mutex);
+ else
+ au_hdir_unlock(h_dir, dir, bindex);
+
+ AuTraceErr(err);
+ return err;
+}
+
+/* ---------------------------------------------------------------------- */
+
+/*
+ * returns a newly allocated branch. @new_nbranch is a number of branches
+ * after adding a branch.
+ */
+static struct au_branch *alloc_addbr(struct super_block *sb, int new_nbranch)
+{
+ struct au_branch **branchp, *add_branch;
+ int sz;
+ void *p;
+ struct dentry *root;
+ struct inode *inode;
+ struct au_hinode *hinodep;
+ struct au_hdentry *hdentryp;
+
+ LKTRTrace("new_nbranch %d\n", new_nbranch);
+ SiMustWriteLock(sb);
+ root = sb->s_root;
+ DiMustWriteLock(root);
+ inode = root->d_inode;
+ IiMustWriteLock(inode);
+
+ add_branch = kmalloc(sizeof(*add_branch), GFP_KERNEL);
+ if (unlikely(!add_branch))
+ goto out;
+
+ sz = sizeof(*branchp) * (new_nbranch - 1);
+ if (unlikely(!sz))
+ sz = sizeof(*branchp);
+ p = au_sbi(sb)->si_branch;
+ branchp = au_kzrealloc(p, sz, sizeof(*branchp) * new_nbranch,
+ GFP_KERNEL);
+ if (unlikely(!branchp))
+ goto out_br;
+ au_sbi(sb)->si_branch = branchp;
+
+ sz = sizeof(*hdentryp) * (new_nbranch - 1);
+ if (unlikely(!sz))
+ sz = sizeof(*hdentryp);
+ p = au_di(root)->di_hdentry;
+ hdentryp = au_kzrealloc(p, sz, sizeof(*hdentryp) * new_nbranch,
+ GFP_KERNEL);
+ if (unlikely(!hdentryp))
+ goto out_br;
+ au_di(root)->di_hdentry = hdentryp;
+
+ sz = sizeof(*hinodep) * (new_nbranch - 1);
+ if (unlikely(!sz))
+ sz = sizeof(*hinodep);
+ p = au_ii(inode)->ii_hinode;
+ hinodep = au_kzrealloc(p, sz, sizeof(*hinodep) * new_nbranch,
+ GFP_KERNEL);
+ if (unlikely(!hinodep))
+ goto out_br;
+ au_ii(inode)->ii_hinode = hinodep;
+ return add_branch; /* success */
+
+ out_br:
+ kfree(add_branch);
+ out:
+ AuTraceErr(-ENOMEM);
+ return ERR_PTR(-ENOMEM);
+}
+
+/*
+ * test if the branch permission is legal or not.
+ */
+static int test_br(struct super_block *sb, struct inode *inode, int brperm,
+ char *path)
+{
+ int err;
+
+ err = 0;
+ if (unlikely(au_br_writable(brperm) && IS_RDONLY(inode))) {
+ AuErr("write permission for readonly fs or inode, %s\n", path);
+ err = -EINVAL;
+ }
+
+ AuTraceErr(err);
+ return err;
+}
+
+/*
+ * returns:
+ * 0: success, the caller will add it
+ * plus: success, it is already unified, the caller should ignore it
+ * minus: error
+ */
+static int test_add(struct super_block *sb, struct au_opt_add *add, int remount)
+{
+ int err;
+ struct dentry *root;
+ struct inode *inode, *h_inode;
+ aufs_bindex_t bend, bindex;
+
+ LKTRTrace("%s, remo%d\n", add->path, remount);
+
+ root = sb->s_root;
+ bend = au_sbend(sb);
+ if (unlikely(bend >= 0
+ && au_find_dbindex(root, add->nd.path.dentry) >= 0)) {
+ err = 1;
+ if (!remount) {
+ err = -EINVAL;
+ AuErr("%s duplicated\n", add->path);
+ }
+ goto out;
+ }
+
+ err = -ENOSPC; /* -E2BIG; */
+ if (unlikely(AUFS_BRANCH_MAX <= add->bindex
+ || AUFS_BRANCH_MAX - 1 <= bend)) {
+ AuErr("number of branches exceeded %s\n", add->path);
+ goto out;
+ }
+
+ err = -EDOM;
+ if (unlikely(add->bindex < 0 || bend + 1 < add->bindex)) {
+ AuErr("bad index %d\n", add->bindex);
+ goto out;
+ }
+
+ inode = add->nd.path.dentry->d_inode;
+ AuDebugOn(!inode || !S_ISDIR(inode->i_mode));
+ err = -ENOENT;
+ if (unlikely(!inode->i_nlink)) {
+ AuErr("no existence %s\n", add->path);
+ goto out;
+ }
+
+ err = -EINVAL;
+ if (unlikely(inode->i_sb == sb)) {
+ AuErr("%s must be outside\n", add->path);
+ goto out;
+ }
+
+ if (unlikely(au_test_nested(inode->i_sb))) {
+ AuErr("nested " AUFS_NAME " %s\n", add->path);
+ goto out;
+ }
+
+ if (unlikely(!strcmp(au_sbtype(inode->i_sb), "unionfs"))) {
+ AuErr("unsupported filesystem, %s\n", add->path);
+ goto out;
+ }
+
+ if (unlikely(au_test_unsupported_nfs(inode->i_sb))) {
+ AuErr(AuNoNfsBranchMsg " %s\n", add->path);
+ goto out;
+ }
+
+ err = test_br(sb, add->nd.path.dentry->d_inode, add->perm, add->path);
+ if (unlikely(err))
+ goto out;
+
+ if (bend < 0)
+ return 0; /* success */
+
+ h_inode = au_h_dptr(root, 0)->d_inode;
+ if (unlikely(au_opt_test(au_mntflags(sb), WARN_PERM)
+ && ((h_inode->i_mode & S_IALLUGO)
+ != (inode->i_mode & S_IALLUGO)
+ || h_inode->i_uid != inode->i_uid
+ || h_inode->i_gid != inode->i_gid)))
+ AuWarn("uid/gid/perm %s %u/%u/0%o, %u/%u/0%o\n",
+ add->path,
+ inode->i_uid, inode->i_gid, (inode->i_mode & S_IALLUGO),
+ h_inode->i_uid, h_inode->i_gid,
+ (h_inode->i_mode & S_IALLUGO));
+
+ err = -EINVAL;
+ for (bindex = 0; bindex <= bend; bindex++)
+ if (unlikely(test_overlap(sb, add->nd.path.dentry,
+ au_h_dptr(root, bindex)))) {
+ AuErr("%s is overlapped\n", add->path);
+ goto out;
+ }
+ err = 0;
+
+ out:
+ AuTraceErr(err);
+ return err;
+}
+
+static int au_br_init(struct au_branch *br, struct super_block *sb,
+ struct au_opt_add *add)
+{
+ int err;
+
+ AuTraceEnter();
+
+ err = 0;
+ au_rw_init_nolock(&br->br_wh_rwsem);
+ br->br_plink = NULL;
+ br->br_wh = NULL;
+ if (unlikely(au_br_writable(add->perm))) {
+ err = init_br_wh(sb, /*bindex*/-1, br, add->perm,
+ add->nd.path.dentry, add->nd.path.mnt);
+ if (unlikely(err))
+ goto out;
+ }
+
+ br->br_xino = NULL;
+ br->br_mnt = mntget(add->nd.path.mnt);
+ if (au_opt_test(au_mntflags(sb), XINO)) {
+ err = au_xino_br(sb, br, add->nd.path.dentry->d_inode->i_ino,
+ au_sbr(sb, 0)->br_xino, /*do_test*/1);
+ if (unlikely(err)) {
+ AuDebugOn(br->br_xino);
+ goto out;
+ }
+ }
+
+ atomic_set(&br->br_wh_running, 0);
+ br->br_id = au_new_br_id(sb);
+ br->br_perm = add->perm;
+ atomic_set(&br->br_count, 0);
+ br->br_bytes = 0;
+ br->br_xino_upper = AUFS_XINO_TRUNC_INIT;
+ atomic_set(&br->br_xino_running, 0);
+ sysaufs_br_init(br);
+ br->br_generation = au_sigen(sb);
+ /* smp_mb(); */ /* atomic_set */
+
+ out:
+ AuTraceErr(err);
+ return err;
+}
+
+int au_br_add(struct super_block *sb, struct au_opt_add *add, int remount)
+{
+ int err, amount;
+ aufs_bindex_t bend, add_bindex;
+ struct dentry *root, *dentry;
+ struct au_iinfo *iinfo;
+ struct au_sbinfo *sbinfo;
+ struct au_dinfo *dinfo;
+ struct inode *root_inode, *inode;
+ unsigned long long maxb;
+ struct au_branch **branchp, *add_branch;
+ struct au_hdentry *hdentryp;
+ struct au_hinode *hinodep;
+
+ dentry = add->nd.path.dentry;
+ LKTRTrace("b%d, %s, 0x%x, %.*s\n",
+ add->bindex, add->path, add->perm, AuDLNPair(dentry));
+ SiMustWriteLock(sb);
+ root = sb->s_root;
+ DiMustWriteLock(root);
+ root_inode = root->d_inode;
+ IMustLock(root_inode);
+ IiMustWriteLock(root_inode);
+
+ err = test_add(sb, add, remount);
+ if (unlikely(err < 0))
+ goto out;
+ if (err)
+ return 0; /* success */
+
+ bend = au_sbend(sb);
+ add_branch = alloc_addbr(sb, bend + 2);
+ err = PTR_ERR(add_branch);
+ if (IS_ERR(add_branch))
+ goto out;
+ err = au_br_init(add_branch, sb, add);
+ if (unlikely(err)) {
+ kfree(add_branch);
+ goto out;
+ }
+
+ add_bindex = add->bindex;
+ if (remount)
+ sysaufs_brs_del(sb, add_bindex);
+
+ sbinfo = au_sbi(sb);
+ dinfo = au_di(root);
+ iinfo = au_ii(root_inode);
+
+ amount = bend + 1 - add_bindex;
+ branchp = sbinfo->si_branch + add_bindex;
+ memmove(branchp + 1, branchp, sizeof(*branchp) * amount);
+ *branchp = add_branch;
+ hdentryp = dinfo->di_hdentry + add_bindex;
+ memmove(hdentryp + 1, hdentryp, sizeof(*hdentryp) * amount);
+ au_h_dentry_init(hdentryp);
+ hinodep = iinfo->ii_hinode + add_bindex;
+ memmove(hinodep + 1, hinodep, sizeof(*hinodep) * amount);
+ hinodep->hi_inode = NULL;
+ au_hin_init(hinodep, NULL);
+
+ sbinfo->si_bend++;
+ dinfo->di_bend++;
+ iinfo->ii_bend++;
+ if (unlikely(bend < 0)) {
+ sbinfo->si_bend = 0;
+ dinfo->di_bstart = 0;
+ iinfo->ii_bstart = 0;
+ }
+ inode = dentry->d_inode;
+ au_set_h_dptr(root, add_bindex, dget(dentry));
+ au_set_h_iptr(root_inode, add_bindex, igrab(inode), 0);
+ if (remount)
+ sysaufs_brs_add(sb, add_bindex);
+
+ if (!add_bindex)
+ au_cpup_attr_all(root_inode);
+ else
+ au_add_nlink(root_inode, inode);
+ maxb = dentry->d_sb->s_maxbytes;
+ if (sb->s_maxbytes < maxb)
+ sb->s_maxbytes = maxb;
+
+ out:
+ AuTraceErr(err);
+ return err;
+}
+
+/* ---------------------------------------------------------------------- */
+
+#define AuVerbose(do_info, fmt, args...) do { \
+ if (!do_info) \
+ LKTRTrace(fmt, ##args); \
+ else \
+ AuInfo(fmt, ##args); \
+} while (0)
+
+/*
+ * test if the branch is deletable or not.
+ */
+static int test_dentry_busy(struct dentry *root, aufs_bindex_t bindex,
+ au_gen_t sigen)
+{
+ int err, i, j, ndentry, verbose;
+ struct au_dcsub_pages dpages;
+ struct au_dpage *dpage;
+ struct dentry *d;
+ aufs_bindex_t bstart, bend;
+ struct inode *inode;
+
+ LKTRTrace("b%d, gen%d\n", bindex, sigen);
+ SiMustWriteLock(root->d_sb);
+
+ err = au_dpages_init(&dpages, GFP_TEMPORARY);
+ if (unlikely(err))
+ goto out;
+ err = au_dcsub_pages(&dpages, root, NULL, NULL);
+ if (unlikely(err))
+ goto out_dpages;
+
+ verbose = !!au_opt_test(au_mntflags(root->d_sb), VERBOSE);
+ for (i = 0; !err && i < dpages.ndpage; i++) {
+ dpage = dpages.dpages + i;
+ ndentry = dpage->ndentry;
+ for (j = 0; !err && j < ndentry; j++) {
+ d = dpage->dentries[j];
+ AuDebugOn(!atomic_read(&d->d_count));
+ inode = d->d_inode;
+ AuDebugOn(!inode);
+ if (au_digen(d) == sigen
+ && au_iigen(inode) == sigen)
+ di_read_lock_child(d, AuLock_IR);
+ else {
+ di_write_lock_child(d);
+ err = au_reval_dpath(d, sigen);
+ if (!err)
+ di_downgrade_lock(d, AuLock_IR);
+ else {
+ di_write_unlock(d);
+ break;
+ }
+ }
+
+ bstart = au_dbstart(d);
+ bend = au_dbend(d);
+ if (bstart <= bindex
+ && bindex <= bend
+ && au_h_dptr(d, bindex)
+ && (!S_ISDIR(d->d_inode->i_mode)
+ || bstart == bend)) {
+ err = -EBUSY;
+ AuVerbose(verbose, "busy %.*s\n", AuDLNPair(d));
+ }
+ di_read_unlock(d, AuLock_IR);
+ }
+ }
+
+ out_dpages:
+ au_dpages_free(&dpages);
+ out:
+ AuTraceErr(err);
+ return err;
+}
+
+static int test_inode_busy(struct super_block *sb, aufs_bindex_t bindex,
+ au_gen_t sigen)
+{
+ int err, verbose;
+ struct inode *i;
+ aufs_bindex_t bstart, bend;
+
+ LKTRTrace("b%d, gen%d\n", bindex, sigen);
+ SiMustWriteLock(sb);
+
+ err = 0;
+ verbose = !!au_opt_test(au_mntflags(sb), VERBOSE);
+ list_for_each_entry(i, &sb->s_inodes, i_sb_list) {
+ AuDebugOn(!atomic_read(&i->i_count));
+ if (!list_empty(&i->i_dentry))
+ continue;
+
+ if (au_iigen(i) == sigen)
+ ii_read_lock_child(i);
+ else {
+ ii_write_lock_child(i);
+ err = au_refresh_hinode_self(i);
+ if (!err)
+ ii_downgrade_lock(i);
+ else {
+ ii_write_unlock(i);
+ break;
+ }
+ }
+
+ bstart = au_ibstart(i);
+ bend = au_ibend(i);
+ if (bstart <= bindex
+ && bindex <= bend
+ && au_h_iptr(i, bindex)
+ && (!S_ISDIR(i->i_mode) || bstart == bend)) {
+ err = -EBUSY;
+ AuVerbose(verbose, "busy i%lu\n", i->i_ino);
+ ii_read_unlock(i);
+ break;
+ }
+ ii_read_unlock(i);
+ }
+
+ AuTraceErr(err);
+ return err;
+}
+
+static int test_children_busy(struct dentry *root, aufs_bindex_t bindex)
+{
+ int err;
+ au_gen_t sigen;
+
+ LKTRTrace("b%d\n", bindex);
+ SiMustWriteLock(root->d_sb);
+ DiMustWriteLock(root);
+ /* dont trust BKL */
+ AuDebugOn(!kernel_locked());
+
+ sigen = au_sigen(root->d_sb);
+ DiMustNoWaiters(root);
+ IiMustNoWaiters(root->d_inode);
+ di_write_unlock(root);
+ err = test_dentry_busy(root, bindex, sigen);
+ if (!err)
+ err = test_inode_busy(root->d_sb, bindex, sigen);
+ di_write_lock_child(root); /* aufs_write_lock() calls ..._child() */
+
+ AuTraceErr(err);
+ return err;
+}
+
+int au_br_del(struct super_block *sb, struct au_opt_del *del, int remount)
+{
+ int err, do_wh, rerr, verbose;
+ struct dentry *root;
+ struct inode *inode, *hidden_dir;
+ aufs_bindex_t bindex, bend, br_id;
+ struct au_sbinfo *sbinfo;
+ struct au_dinfo *dinfo;
+ struct au_iinfo *iinfo;
+ struct au_branch *br;
+ unsigned int mnt_flags;
+
+ LKTRTrace("%s, %.*s\n", del->path, AuDLNPair(del->h_root));
+ SiMustWriteLock(sb);
+ root = sb->s_root;
+ DiMustWriteLock(root);
+ inode = root->d_inode;
+ IiMustWriteLock(inode);
+
+ err = 0;
+ bindex = au_find_dbindex(root, del->h_root);
+ if (bindex < 0) {
+ if (remount)
+ goto out; /* success */
+ err = -ENOENT;
+ AuErr("%s no such branch\n", del->path);
+ goto out;
+ }
+ LKTRTrace("bindex b%d\n", bindex);
+
+ err = -EBUSY;
+ mnt_flags = au_mntflags(sb);
+ verbose = au_opt_test(mnt_flags, VERBOSE);
+ bend = au_sbend(sb);
+ if (unlikely(!bend)) {
+ AuVerbose(verbose, "no more branches left\n");
+ goto out;
+ }
+ br = au_sbr(sb, bindex);
+ if (unlikely(au_br_count(br))) {
+ AuVerbose(verbose, "%d file(s) opened\n", au_br_count(br));
+ goto out;
+ }
+
+ do_wh = 0;
+ hidden_dir = del->h_root->d_inode;
+ if (br->br_wh || br->br_plink) {
+#if 0 /* reserved for future use */
+ /* remove whiteout base */
+ err = init_br_wh(sb, bindex, br, AuBr_RO, del->h_root,
+ br->br_mnt);
+ if (unlikely(err))
+ goto out;
+#else
+ dput(br->br_wh);
+ dput(br->br_plink);
+ br->br_plink = NULL;
+ br->br_wh = NULL;
+#endif
+ do_wh = 1;
+ }
+
+ err = test_children_busy(root, bindex);
+ if (unlikely(err)) {
+ if (unlikely(do_wh))
+ goto out_wh;
+ goto out;
+ }
+
+ err = 0;
+ if (remount)
+ sysaufs_brs_del(sb, bindex);
+ sbinfo = au_sbi(sb);
+ dinfo = au_di(root);
+ iinfo = au_ii(inode);
+
+ dput(au_h_dptr(root, bindex));
+ au_hiput(iinfo->ii_hinode + bindex);
+ br_id = br->br_id;
+ free_branch(br);
+
+ /* todo: realloc and shrink memory? */
+ if (bindex < bend) {
+ const aufs_bindex_t n = bend - bindex;
+ struct au_branch **brp;
+ struct au_hdentry *hdp;
+ struct au_hinode *hip;
+
+ brp = sbinfo->si_branch + bindex;
+ memmove(brp, brp + 1, sizeof(*brp) * n);
+ hdp = dinfo->di_hdentry + bindex;
+ memmove(hdp, hdp + 1, sizeof(*hdp) * n);
+ hip = iinfo->ii_hinode + bindex;
+ memmove(hip, hip + 1, sizeof(*hip) * n);
+ }
+ sbinfo->si_branch[0 + bend] = NULL;
+ dinfo->di_hdentry[0 + bend].hd_dentry = NULL;
+ iinfo->ii_hinode[0 + bend].hi_inode = NULL;
+ au_hin_init(iinfo->ii_hinode + bend, NULL);
+
+ sbinfo->si_bend--;
+ dinfo->di_bend--;
+ iinfo->ii_bend--;
+ if (remount)
+ sysaufs_brs_add(sb, bindex);
+
+ if (!bindex)
+ au_cpup_attr_all(inode);
+ else
+ au_sub_nlink(inode, del->h_root->d_inode);
+ if (au_opt_test(mnt_flags, PLINK))
+ au_plink_half_refresh(sb, br_id);
+
+ if (sb->s_maxbytes == del->h_root->d_sb->s_maxbytes) {
+ bend--;
+ sb->s_maxbytes = 0;
+ for (bindex = 0; bindex <= bend; bindex++) {
+ unsigned long long maxb;
+ maxb = au_sbr_sb(sb, bindex)->s_maxbytes;
+ if (sb->s_maxbytes < maxb)
+ sb->s_maxbytes = maxb;
+ }
+ }
+ goto out; /* success */
+
+ out_wh:
+ /* revert */
+ rerr = init_br_wh(sb, bindex, br, br->br_perm, del->h_root, br->br_mnt);
+ if (rerr)
+ AuWarn("failed re-creating base whiteout, %s. (%d)\n",
+ del->path, rerr);
+ out:
+ AuTraceErr(err);
+ return err;
+}
+
+static int do_need_sigen_inc(int a, int b)
+{
+ return (au_br_whable(a) && !au_br_whable(b));
+}
+
+static int need_sigen_inc(int old, int new)
+{
+ return (do_need_sigen_inc(old, new)
+ || do_need_sigen_inc(new, old));
+}
+
+int au_br_mod(struct super_block *sb, struct au_opt_mod *mod, int remount,
+ int *do_update)
+{
+ int err;
+ struct dentry *root;
+ aufs_bindex_t bindex;
+ struct au_branch *br;
+ struct inode *hidden_dir;
+
+ LKTRTrace("%s, %.*s, 0x%x\n",
+ mod->path, AuDLNPair(mod->h_root), mod->perm);
+ SiMustWriteLock(sb);
+ root = sb->s_root;
+ DiMustWriteLock(root);
+ IiMustWriteLock(root->d_inode);
+
+ bindex = au_find_dbindex(root, mod->h_root);
+ if (bindex < 0) {
+ if (remount)
+ return 0; /* success */
+ err = -ENOENT;
+ AuErr("%s no such branch\n", mod->path);
+ goto out;
+ }
+ LKTRTrace("bindex b%d\n", bindex);
+
+ hidden_dir = mod->h_root->d_inode;
+ err = test_br(sb, hidden_dir, mod->perm, mod->path);
+ if (unlikely(err))
+ goto out;
+
+ br = au_sbr(sb, bindex);
+ if (br->br_perm == mod->perm)
+ return 0; /* success */
+
+ if (au_br_writable(br->br_perm)) {
+#if 1
+ /* remove whiteout base */
+ /* todo: mod->perm? */
+ err = init_br_wh(sb, bindex, br, AuBr_RO, mod->h_root,
+ br->br_mnt);
+ if (unlikely(err))
+ goto out;
+#else /* reserved for future use */
+ dput(br->br_wh);
+ dput(br->br_plink);
+ br->br_plink = NULL;
+ br->br_wh = NULL;
+#endif
+
+ if (!au_br_writable(mod->perm)) {
+ /* rw --> ro, file might be mmapped */
+ struct file *file, *hf;
+
+#if 1 /* todo: test more? */
+ DiMustNoWaiters(root);
+ IiMustNoWaiters(root->d_inode);
+ di_write_unlock(root);
+
+ /*
+ * no need file_list_lock()
+ * since BKL (and sbinfo) is locked
+ */
+ AuDebugOn(!kernel_locked());
+ list_for_each_entry(file, &sb->s_files, f_u.fu_list) {
+ LKTRTrace("%.*s\n", AuDLNPair(file->f_dentry));
+ if (!au_test_aufs_file(file))
+ continue;
+
+ fi_read_lock(file);
+ if (!S_ISREG(file->f_dentry->d_inode->i_mode)
+ || !(file->f_mode & FMODE_WRITE)
+ || au_fbstart(file) != bindex) {
+ FiMustNoWaiters(file);
+ fi_read_unlock(file);
+ continue;
+ }
+
+ if (unlikely(au_test_mmapped(file))) {
+ err = -EBUSY;
+ FiMustNoWaiters(file);
+ fi_read_unlock(file);
+ break;
+ }
+
+ /* todo: already flushed? */
+ hf = au_h_fptr(file, au_fbstart(file));
+ hf->f_flags = au_file_roflags(hf->f_flags);
+ hf->f_mode &= ~FMODE_WRITE;
+ put_write_access(hf->f_dentry->d_inode);
+ FiMustNoWaiters(file);
+ fi_read_unlock(file);
+ }
+
+ /* aufs_write_lock() calls ..._child() */
+ di_write_lock_child(root);
+#endif
+ }
+ }
+
+ if (!err) {
+ *do_update |= need_sigen_inc(br->br_perm, mod->perm);
+ br->br_perm = mod->perm;
+ }
+
+ out:
+ AuTraceErr(err);
+ return err;
+}
diff --git a/fs/aufs/branch.h b/fs/aufs/branch.h
new file mode 100644
index 0000000..6e4a9aa
--- /dev/null
+++ b/fs/aufs/branch.h
@@ -0,0 +1,352 @@
+/*
+ * Copyright (C) 2005-2008 Junjiro Okajima
+ *
+ * This program, aufs is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+/*
+ * branch filesystems and xino for them
+ */
+
+#ifndef __AUFS_BRANCH_H__
+#define __AUFS_BRANCH_H__
+
+#ifdef __KERNEL__
+
+#include <linux/fs.h>
+#include <linux/mount.h>
+#include <linux/sysfs.h>
+#include <linux/aufs_type.h>
+#include "misc.h"
+#include "super.h"
+
+/* ---------------------------------------------------------------------- */
+
+/* an entry in a xino file */
+struct au_xino_entry {
+ ino_t ino;
+ /* __u32 h_gen; */ /* reserved for future use */
+} __packed;
+
+/* reserved for future use */
+/* #define AuXino_INVALID_HGEN (-1) */
+
+/* a xino file */
+struct au_xino_file {
+ struct file *xi_file;
+
+ /* reserved for future use */
+#if 0
+ struct file **xi_file;
+
+ /* array management */
+ unsigned long long xi_limit; /* Max xino file size */
+ unsigned long long xi_size; /* s_maxbytes */
+
+ /* truncation */
+ u64 xi_upper; /* watermark in bytes */
+ u64 xi_step; /* to next watermark in bytes */
+#endif
+
+ /* truncation */
+ blkcnt_t xi_upper; /* watermark in blocks */
+ atomic_t xi_running;
+};
+
+/* protected by superblock rwsem */
+struct au_branch {
+ struct file *br_xino;
+#if 0 /* reserved for future use */
+ struct au_xino_file *br_xino;
+#endif
+
+ aufs_bindex_t br_id;
+
+ int br_perm;
+ struct vfsmount *br_mnt;
+ atomic_t br_count;
+
+ /* whiteout base */
+ struct au_rwsem br_wh_rwsem;
+ struct dentry *br_wh;
+ atomic_t br_wh_running;
+
+ /* pseudo-link dir */
+ struct dentry *br_plink;
+
+#if 1 /* reserved for future use */
+ /* xino truncation */
+ blkcnt_t br_xino_upper; /* watermark in blocks */
+ atomic_t br_xino_running;
+#endif
+
+ /* mfs mode */
+ u64 br_bytes;
+
+#ifdef CONFIG_SYSFS
+ /* an entry under sysfs per mount-point */
+ char br_name[8];
+ struct attribute br_attr;
+#endif
+
+ au_gen_t br_generation;
+};
+
+/* ---------------------------------------------------------------------- */
+
+/* branch permission and attribute */
+enum {
+ AuBr_RW, /* writable, linkable wh */
+ AuBr_RO, /* readonly, no wh */
+ AuBr_RR, /* natively readonly, no wh */
+
+ AuBr_RWNoLinkWH, /* un-linkable whiteouts */
+
+ AuBr_ROWH,
+ AuBr_RRWH, /* whiteout-able */
+
+ AuBr_Last
+};
+
+static inline int au_br_writable(int brperm)
+{
+ return (brperm == AuBr_RW || brperm == AuBr_RWNoLinkWH);
+}
+
+static inline int au_br_whable(int brperm)
+{
+ return (brperm == AuBr_RW
+ || brperm == AuBr_ROWH
+ || brperm == AuBr_RRWH);
+}
+
+#if 0 /* reserved for future use */
+static inline int au_br_linkable_wh(int brperm)
+{
+ return (brperm == AuBr_RW);
+}
+#endif
+
+static inline int au_br_hinotifyable(int brperm)
+{
+#ifdef CONFIG_AUFS_HINOTIFY
+ return (brperm != AuBr_RR && brperm != AuBr_RRWH);
+#else
+ return 0;
+#endif
+}
+
+/* ---------------------------------------------------------------------- */
+
+/* branch.c */
+struct au_sbinfo;
+void au_br_free(struct au_sbinfo *sinfo);
+int au_test_def_rr(struct super_block *h_sb);
+int au_br_index(struct super_block *sb, aufs_bindex_t br_id);
+struct au_opt_add;
+int au_br_add(struct super_block *sb, struct au_opt_add *add, int remount);
+struct au_opt_del;
+int au_br_del(struct super_block *sb, struct au_opt_del *del, int remount);
+struct au_opt_mod;
+int au_br_mod(struct super_block *sb, struct au_opt_mod *mod, int remount,
+ int *do_update);
+
+/* xino.c */
+int au_xib_trunc(struct super_block *sb);
+
+struct file *au_xino_create(struct super_block *sb, char *fname, int silent,
+ struct dentry *parent);
+ino_t au_xino_new_ino(struct super_block *sb);
+int au_xino_write0(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino,
+ ino_t ino);
+int au_xino_write(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino,
+ struct au_xino_entry *xinoe);
+int au_xino_read(struct super_block *sb, aufs_bindex_t bindex, ino_t h_ino,
+ struct au_xino_entry *xinoe);
+int au_xino_br(struct super_block *sb, struct au_branch *br, ino_t hino,
+ struct file *base_file, int do_test);
+int au_xino_trunc(struct super_block *sb, aufs_bindex_t bindex);
+
+struct au_opt_xino;
+int au_xino_set(struct super_block *sb, struct au_opt_xino *xino, int remount);
+void au_xino_clr(struct super_block *sb);
+struct file *au_xino_def(struct super_block *sb);
+
+/* ---------------------------------------------------------------------- */
+
+/* todo: memory barrier? */
+static inline int au_br_count(struct au_branch *br)
+{
+ return atomic_read(&br->br_count);
+}
+
+static inline int au_br_get(struct au_branch *br)
+{
+ return atomic_inc_return(&br->br_count);
+}
+
+static inline int au_br_put(struct au_branch *br)
+{
+ return atomic_dec_return(&br->br_count);
+}
+
+static inline au_gen_t au_br_gen(struct au_branch *br)
+{
+ return br->br_generation;
+}
+
+/*
+ * test if the @br is readonly or not.
+ */
+static inline int au_br_rdonly(struct au_branch *br)
+{
+ return ((br->br_mnt->mnt_sb->s_flags & MS_RDONLY)
+ || !au_br_writable(br->br_perm))
+ ? -EROFS : 0;
+}
+
+/* ---------------------------------------------------------------------- */
+
+/* Superblock to branch */
+static inline
+aufs_bindex_t au_sbr_id(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_sbr(sb, bindex)->br_id;
+}
+
+static inline
+struct vfsmount *au_sbr_mnt(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_sbr(sb, bindex)->br_mnt;
+}
+
+static inline
+struct super_block *au_sbr_sb(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_sbr_mnt(sb, bindex)->mnt_sb;
+}
+
+#if 0 /* reserved for future use */
+static inline int au_sbr_count(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_br_count(au_sbr(sb, bindex));
+}
+
+static inline void au_sbr_get(struct super_block *sb, aufs_bindex_t bindex)
+{
+ au_br_get(au_sbr(sb, bindex));
+}
+#endif
+
+static inline void au_sbr_put(struct super_block *sb, aufs_bindex_t bindex)
+{
+ au_br_put(au_sbr(sb, bindex));
+}
+
+static inline int au_sbr_perm(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_sbr(sb, bindex)->br_perm;
+}
+
+static inline int au_sbr_whable(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_br_whable(au_sbr_perm(sb, bindex));
+}
+
+static inline int au_test_trunc_xino(struct super_block *sb)
+{
+ return au_test_tmpfs(sb);
+}
+
+/* temporary support for i#1 in cramfs */
+static inline int au_test_unique_ino(struct dentry *h_dentry, ino_t h_ino)
+{
+#if defined(CONFIG_CRAMFS) || defined(CONFIG_CRAMFS_MODULE)
+ if (unlikely(h_dentry->d_sb->s_magic == CRAMFS_MAGIC))
+ return (h_ino != 1);
+#endif
+ return 1;
+}
+
+#ifdef CONFIG_AUFS_BR_NFS
+static inline int au_test_unsupported_nfs(struct super_block *h_sb)
+{
+ return 0;
+}
+
+static inline struct vfsmount *au_do_nfsmnt(struct vfsmount *h_mnt)
+{
+ if (!au_test_nfs(h_mnt->mnt_sb))
+ return NULL;
+ return h_mnt;
+}
+
+/* it doesn't mntget() */
+static inline
+struct vfsmount *au_nfsmnt(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return au_do_nfsmnt(au_sbr_mnt(sb, bindex));
+}
+
+#define AuNoNfsBranchMsg "dummy"
+
+#else
+static inline int au_test_unsupported_nfs(struct super_block *h_sb)
+{
+ return au_test_nfs(h_sb);
+}
+
+static inline struct vfsmount *au_do_nfsmnt(struct vfsmount *h_mnt)
+{
+ return NULL;
+}
+
+static inline
+struct vfsmount *au_nfsmnt(struct super_block *sb, aufs_bindex_t bindex)
+{
+ return NULL;
+}
+
+#define AuNoNfsBranchMsg "NFS branch is not supported" \
+ ", try some configurations and patches included in aufs source CVS."
+
+#endif /* CONFIG_AUFS_BR_NFS */
+
+/* ---------------------------------------------------------------------- */
+
+/*
+ * br_wh_read_lock, br_wh_write_lock
+ * br_wh_read_unlock, br_wh_write_unlock, br_wh_downgrade_lock
+ */
+AuSimpleRwsemFuncs(br_wh, struct au_branch *br, br->br_wh_rwsem);
+
+/* to debug easier, do not make them inlined functions */
+#define BrWhMustReadLock(br) do { \
+ /* SiMustAnyLock(sb); */ \
+ AuRwMustReadLock(&(br)->br_wh_rwsem); \
+} while (0)
+
+#define BrWhMustWriteLock(br) do { \
+ /* SiMustAnyLock(sb); */ \
+ AuRwMustWriteLock(&(br)->br_wh_rwsem); \
+} while (0)
+
+#define BrWhMustAnyLock(br) do { \
+ /* SiMustAnyLock(sb); */ \
+ AuRwMustAnyLock(&(br)->br_wh_rwsem); \
+} while (0)
+
+#endif /* __KERNEL__ */
+#endif /* __AUFS_BRANCH_H__ */
--
1.5.5.1.308.g1fbb5.dirty

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/