Message-Id: <1210948402629-git-send-email-hooanon05@yahoo.co.jp>
Date:	Fri, 16 May 2008 23:33:03 +0900
From:	hooanon05@...oo.co.jp
To:	linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Cc:	Junjiro Okajima <hooanon05@...oo.co.jp>
Subject: [PATCH 49/67] aufs inode operations (del entry)

From: Junjiro Okajima <hooanon05@...oo.co.jp>

	initial commit
	aufs inode operations (del entry)

Signed-off-by: Junjiro Okajima <hooanon05@...oo.co.jp>
---
 fs/aufs/i_op_del.c |  582 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 582 insertions(+), 0 deletions(-)

diff --git a/fs/aufs/i_op_del.c b/fs/aufs/i_op_del.c
new file mode 100644
index 0000000..5b84cca
--- /dev/null
+++ b/fs/aufs/i_op_del.c
@@ -0,0 +1,582 @@
+/*
+ * 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
+ */
+
+/*
+ * inode operations (del entry)
+ *
+ * $Id: i_op_del.c,v 1.4 2008/05/04 23:54:53 sfjro Exp $
+ */
+
+#include "aufs.h"
+
+/* returns,
+ * 0: wh is unnecessary
+ * plus: wh is necessary
+ * minus: error
+ */
+int au_wr_dir_need_wh(struct dentry *dentry, int isdir, aufs_bindex_t *bcpup,
+		      struct dentry *locked)
+{
+	int need_wh, err;
+	aufs_bindex_t bstart;
+	struct dentry *h_dentry;
+	struct super_block *sb;
+
+	LKTRTrace("%.*s, isdir %d, *bcpup %d, locked %p\n",
+		  AuDLNPair(dentry), isdir, *bcpup, locked);
+	sb = dentry->d_sb;
+
+	bstart = au_dbstart(dentry);
+	LKTRTrace("bcpup %d, bstart %d\n", *bcpup, bstart);
+	h_dentry = au_h_dptr(dentry, bstart);
+	if (*bcpup < 0) {
+		*bcpup = bstart;
+		if (au_test_ro(sb, bstart, dentry->d_inode)) {
+			err = AuWbrCopyup(au_sbi(sb), dentry);
+			*bcpup = err;
+			//err = -1;
+			if (unlikely(err < 0))
+				goto out;
+		}
+	} else
+		AuDebugOn(bstart < *bcpup
+			  || au_test_ro(sb, *bcpup, dentry->d_inode));
+	LKTRTrace("bcpup %d, bstart %d\n", *bcpup, bstart);
+
+	if (*bcpup != bstart) {
+		err = au_cpup_dirs(dentry, *bcpup, locked);
+		//err = -1;
+		if (unlikely(err))
+			goto out;
+		need_wh = 1;
+	} else {
+		//struct nameidata nd;
+		aufs_bindex_t old_bend, new_bend, bdiropq = -1;
+		old_bend = au_dbend(dentry);
+		if (isdir) {
+			bdiropq = au_dbdiropq(dentry);
+			au_set_dbdiropq(dentry, -1);
+		}
+		need_wh = au_lkup_dentry(dentry, bstart + 1, /*type*/0,
+					 /*nd*/NULL);
+		err = need_wh;
+		//err = -1;
+		if (isdir)
+			au_set_dbdiropq(dentry, bdiropq);
+		if (unlikely(err < 0))
+			goto out;
+		new_bend = au_dbend(dentry);
+		if (!need_wh && old_bend != new_bend) {
+			au_set_h_dptr(dentry, new_bend, NULL);
+			au_set_dbend(dentry, old_bend);
+#if 0 // remove this?
+		} else if (!au_h_dptr(dentry, new_bend)->d_inode) {
+			LKTRTrace("negative\n");
+			au_set_h_dptr(dentry, new_bend, NULL);
+			au_set_dbend(dentry, old_bend);
+			need_wh = 0;
+#endif
+		}
+	}
+	LKTRTrace("need_wh %d\n", need_wh);
+	err = need_wh;
+
+ out:
+	AuTraceErr(err);
+	return err;
+}
+
+/*
+ * simple tests for the removal inode operations.
+ * following the checks in vfs, plus the parent-child relationship.
+ */
+int au_may_del(struct dentry *dentry, aufs_bindex_t bindex,
+	       struct dentry *h_parent, int isdir, struct au_ndx *ndx)
+{
+	int err, exist;
+	struct super_block *sb;
+	struct dentry *h_dentry;
+	struct inode *h_inode;
+	umode_t h_mode;
+
+	LKTRTrace("%.*s/%.*s, b%d, dir %d\n",
+		  AuDLNPair(h_parent), AuDLNPair(dentry), bindex, isdir);
+
+	sb = dentry->d_sb;
+	exist = !!dentry->d_inode;
+	h_dentry = au_h_dptr(dentry, bindex);
+	h_inode = h_dentry->d_inode;
+	if (exist) {
+		err = -ENOENT;
+		if (unlikely(!h_inode || !h_inode->i_nlink))
+			goto out;
+
+		h_mode = h_inode->i_mode;
+		if (!isdir) {
+			err = -EISDIR;
+			if (unlikely(S_ISDIR(h_mode)))
+				goto out;
+		} else if (unlikely(!S_ISDIR(h_mode))) {
+			err = -ENOTDIR;
+			goto out;
+		}
+	} else {
+		/* rename(2) case */
+		err = -EIO;
+		if (unlikely(h_inode))
+			goto out;
+	}
+
+	err = -ENOENT;
+	/* expected parent dir is locked */
+	if (unlikely(h_parent != h_dentry->d_parent))
+		goto out;
+	err = 0;
+
+	/*
+	 * some filesystem may unlink a dir and corrupt its consistency.
+	 * so let's try heavy test.
+	 */
+	if (1 /*unlikely(au_opt_test(au_mntflags(sb), UDBA_INOTIFY))*/) {
+		struct dentry *h_latest;
+		struct qstr *qstr = &dentry->d_name;
+
+		err = -EACCES;
+		if (unlikely(au_test_h_perm(h_parent->d_inode,
+					    MAY_EXEC | MAY_WRITE,
+					    au_ftest_ndx(ndx->flags, DLGT))))
+			goto out;
+
+		h_latest = au_sio_lkup_one(qstr->name, h_parent, qstr->len,
+					   ndx);
+		//err = PTR_ERR(h_latest);
+		err = -EIO;
+		if (IS_ERR(h_latest))
+			goto out;
+		dput(h_latest);
+		if (h_latest == h_dentry)
+			err = 0;
+	}
+
+ out:
+	AuTraceErr(err);
+	return err;
+}
+
+static struct dentry *
+lock_hdir_create_wh(struct dentry *dentry, int isdir, aufs_bindex_t *rbcpup,
+		    struct au_dtime *dt)
+{
+	struct dentry *wh_dentry;
+	int err, need_wh;
+	struct dentry *h_parent, *parent, *gparent;
+	struct inode *dir, *h_dir, *gdir;
+	struct au_ndx ndx;
+	struct super_block *sb;
+	struct au_hinode *hgdir;
+	aufs_bindex_t bcpup;
+	unsigned int mnt_flags;
+
+	LKTRTrace("%.*s, isdir %d\n", AuDLNPair(dentry), isdir);
+
+	need_wh = au_wr_dir_need_wh(dentry, isdir, rbcpup, NULL);
+	err = need_wh;
+	//err = -1;
+	wh_dentry = ERR_PTR(err);
+	if (unlikely(err < 0))
+		goto out;
+
+	//todo: meaningless lock if CONFIG_AUFS_DEBUG is disabled.
+	hgdir = NULL;
+	bcpup = *rbcpup;
+	sb = dentry->d_sb;
+	mnt_flags = au_mntflags(sb);
+	parent = dentry->d_parent; /* dir inode is locked */
+	if (unlikely(au_opt_test(mnt_flags, UDBA_INOTIFY)
+		     && !IS_ROOT(parent))) {
+		gparent = dget_parent(parent);
+		gdir = gparent->d_inode;
+		ii_read_lock_parent2(gdir);
+		hgdir = au_hi(gdir, bcpup);
+		ii_read_unlock(gdir);
+		dput(gparent);
+	}
+	dir = parent->d_inode;
+	IMustLock(dir);
+	h_parent = au_h_dptr(parent, bcpup);
+	h_dir = h_parent->d_inode;
+
+	AuDbgSleep_UdbaRace();
+	au_hdir_lock(h_dir, dir, bcpup);
+	//todo: revalidate the lower dentry?
+
+	if (!au_opt_test(mnt_flags, UDBA_NONE) && au_dbstart(dentry) == bcpup) {
+		ndx.nfsmnt = au_nfsmnt(sb, bcpup);
+		ndx.flags = 0;
+		if (unlikely(au_opt_test_dlgt(mnt_flags)))
+			au_fset_ndx(ndx.flags, DLGT);
+		ndx.nd = NULL;
+		//ndx.br = au_sbr(sb, bcpup);
+		//ndx.nd_file = NULL;
+		err = au_may_del(dentry, bcpup, h_parent, isdir, &ndx);
+		wh_dentry = ERR_PTR(err);
+		if (unlikely(err))
+			goto out_dir;
+	}
+
+	au_dtime_store(dt, parent, h_parent, hgdir);
+	wh_dentry = NULL;
+	if (!need_wh)
+		goto out; /* success, no need to create whiteout */
+
+	ndx.nfsmnt = au_nfsmnt(sb, bcpup);
+	ndx.flags = 0;
+	if (unlikely(au_opt_test_dlgt(mnt_flags)))
+		au_fset_ndx(ndx.flags, DLGT);
+	ndx.nd = NULL;
+	//ndx.br = NULL;
+	wh_dentry = au_wh_create(dir, dentry, bcpup, h_parent, &ndx);
+	//wh_dentry = ERR_PTR(-1);
+	if (!IS_ERR(wh_dentry))
+		goto out; /* success */
+	/* returns with the parent is locked and wh_dentry is DGETed */
+
+ out_dir:
+	au_hdir_unlock(h_dir, dir, bcpup);
+ out:
+	AuTraceErrPtr(wh_dentry);
+	return wh_dentry;
+}
+
+static int renwh_and_rmdir(struct dentry *dentry, aufs_bindex_t bindex,
+			   struct au_nhash *whlist, struct inode *dir)
+{
+	int rmdir_later, err;
+	struct dentry *h_dentry;
+	struct inode *inode, *h_inode;
+	struct super_block *sb;
+
+	LKTRTrace("%.*s, b%d\n", AuDLNPair(dentry), bindex);
+
+	inode = NULL;
+	h_inode = NULL;
+	sb = dentry->d_sb;
+	if (unlikely(au_opt_test(au_mntflags(sb), UDBA_INOTIFY))) {
+		inode = dentry->d_inode;
+		h_inode = au_h_iptr(inode, bindex);
+		au_hdir2_lock(h_inode, inode, bindex);
+	}
+	err = au_whtmp_ren(dir, dentry, bindex, /*noself*/1);
+	if (unlikely(inode))
+		au_hdir_unlock(h_inode, inode, bindex);
+	//err = -1;
+	if (unlikely(err))
+		goto out;
+
+	h_dentry = au_h_dptr(dentry, bindex);
+	if (!au_test_nfs(h_dentry->d_sb)) {
+		const int dirwh = au_sbi(sb)->si_dirwh;
+		rmdir_later = (dirwh <= 1);
+		if (!rmdir_later)
+			rmdir_later = au_nhash_test_longer_wh(whlist, bindex,
+							      dirwh);
+		if (rmdir_later)
+			return rmdir_later;
+	}
+
+	err = au_whtmp_rmdir(h_dentry, whlist, bindex, dir, dentry->d_inode,
+			     /*noself*/1);
+	//err = -1;
+	if (unlikely(err)) {
+		AuIOErr("rmdir %.*s, b%d failed, %d. ignored\n",
+			AuDLNPair(h_dentry), bindex, err);
+		err = 0;
+	}
+
+ out:
+	AuTraceErr(err);
+	return err;
+}
+
+static void epilog(struct inode *dir, struct dentry *dentry,
+		   aufs_bindex_t bindex)
+{
+	//todo: unnecessary?
+	d_drop(dentry);
+	dentry->d_inode->i_ctime = dir->i_ctime;
+
+	if (atomic_read(&dentry->d_count) == 1) {
+		au_set_h_dptr(dentry, au_dbstart(dentry), NULL);
+		au_update_dbstart(dentry);
+	}
+	if (au_ibstart(dir) == bindex)
+		au_cpup_attr_timesizes(dir);
+	dir->i_version++;
+}
+
+/* revert flags */
+#define AuRev_DLGT	1
+#define au_ftest_rev(flags, name)	((flags) & AuRev_##name)
+#define au_fset_rev(flags, name)	{ (flags) |= AuRev_##name; }
+#define au_fclr_rev(flags, name)	{ (flags) &= ~AuRev_##name; }
+#ifndef CONFIG_AUFS_DLGT
+#undef AuRev_DLGT
+#define AuRev_DLGT	0
+#endif
+
+static int do_revert(int err, struct dentry *wh_dentry, struct dentry *dentry,
+		     aufs_bindex_t bwh, struct au_dtime *dt, unsigned int flags)
+{
+	int rerr;
+	struct inode *dir;
+
+	dir = wh_dentry->d_parent->d_inode; /* dir inode is locked */
+	IMustLock(dir);
+	rerr = au_wh_unlink_dentry(dir, wh_dentry, dentry, dir,
+				   au_ftest_rev(flags, DLGT));
+	//rerr = -1;
+	if (!rerr) {
+		au_set_dbwh(dentry, bwh);
+		au_dtime_revert(dt);
+		return 0;
+	}
+
+	AuIOErr("%.*s reverting whiteout failed(%d, %d)\n",
+		AuDLNPair(dentry), err, rerr);
+	return -EIO;
+}
+
+/* ---------------------------------------------------------------------- */
+
+int aufs_unlink(struct inode *dir, struct dentry *dentry)
+{
+	int err, dlgt;
+	struct inode *inode, *h_dir;
+	struct dentry *parent, *wh_dentry, *h_dentry;
+	struct au_dtime dt;
+	aufs_bindex_t bwh, bindex, bstart;
+	struct super_block *sb;
+	struct vfsub_args vargs;
+
+	LKTRTrace("i%lu, %.*s\n", dir->i_ino, AuDLNPair(dentry));
+	IMustLock(dir);
+	inode = dentry->d_inode;
+	if (unlikely(!inode))
+		return -ENOENT; /* possible? */
+	IMustLock(inode);
+
+	aufs_read_lock(dentry, AuLock_DW);
+	parent = dentry->d_parent; /* dir inode is locked */
+	di_write_lock_parent(parent);
+
+	bstart = au_dbstart(dentry);
+	bwh = au_dbwh(dentry);
+	bindex = -1;
+	wh_dentry = lock_hdir_create_wh(dentry, /*isdir*/0, &bindex, &dt);
+	//wh_dentry = ERR_PTR(-1);
+	err = PTR_ERR(wh_dentry);
+	if (IS_ERR(wh_dentry))
+		goto out;
+
+	sb = dir->i_sb;
+	dlgt = !!au_opt_test_dlgt(au_mntflags(sb));
+	AuDebugOn(au_dbstart(dentry) != bstart);
+	h_dentry = au_h_dptr(dentry, bstart);
+	dget(h_dentry);
+
+	if (bindex == bstart) {
+		vfsub_args_init(&vargs, NULL, dlgt, 0);
+		h_dir = h_dentry->d_parent->d_inode; /* dir inode is locked */
+		IMustLock(h_dir);
+		err = vfsub_unlink(h_dir, h_dentry, &vargs);
+		//err = -1;
+	} else {
+		/* dir inode is locked */
+		AuDebugOn(!wh_dentry
+			  || wh_dentry->d_parent != au_h_dptr(parent, bindex));
+		h_dir = wh_dentry->d_parent->d_inode;
+		IMustLock(h_dir);
+		err = 0;
+	}
+
+	if (!err) {
+		drop_nlink(inode);
+#if 0 //todo: update plink
+		if (unlikely(!inode->i_nlink
+			     && au_plink_test(sb, inode)
+			     /* && atomic_read(&inode->i_count) == 2) */)) {
+			au_debug_on();
+			DbgInode(inode);
+			au_debug_off();
+		}
+#endif
+		epilog(dir, dentry, bindex);
+
+		/* update target timestamps */
+		if (bindex == bstart) {
+			au_update_fuse_h_inode(NULL, h_dentry); /*ignore*/
+			inode->i_ctime = h_dentry->d_inode->i_ctime;
+		} else
+			//todo: this timestamp may be reverted later
+			inode->i_ctime = h_dir->i_ctime;
+		goto out_unlock; /* success */
+	}
+
+	/* revert */
+	if (wh_dentry) {
+		int rerr;
+		unsigned int rev_flags;
+
+		rev_flags = 0;
+		if (unlikely(dlgt))
+			au_fset_rev(rev_flags, DLGT);
+		rerr = do_revert(err, wh_dentry, dentry, bwh, &dt, rev_flags);
+		if (rerr)
+			err = rerr;
+	}
+
+ out_unlock:
+	au_hdir_unlock(h_dir, dir, bindex);
+	dput(wh_dentry);
+	dput(h_dentry);
+ out:
+	di_write_unlock(parent);
+	aufs_read_unlock(dentry, AuLock_DW);
+	AuTraceErr(err);
+	return err;
+}
+
+int aufs_rmdir(struct inode *dir, struct dentry *dentry)
+{
+	int err, rmdir_later;
+	struct inode *inode, *h_dir;
+	struct dentry *parent, *wh_dentry, *h_dentry;
+	struct au_dtime dt;
+	aufs_bindex_t bwh, bindex, bstart;
+	struct au_whtmp_rmdir_args *args;
+	struct au_nhash *whlist;
+	struct super_block *sb;
+	unsigned int mnt_flags;
+
+	LKTRTrace("i%lu, %.*s\n", dir->i_ino, AuDLNPair(dentry));
+	IMustLock(dir);
+	inode = dentry->d_inode;
+	if (unlikely(!inode))
+		return -ENOENT; /* possible? */
+	IMustLock(inode);
+
+	whlist = au_nhash_new(GFP_TEMPORARY);
+	err = PTR_ERR(whlist);
+	if (IS_ERR(whlist))
+		goto out;
+
+	err = -ENOMEM;
+	args = kmalloc(sizeof(*args), GFP_TEMPORARY);
+	//args = NULL;
+	if (unlikely(!args))
+		goto out_whlist;
+
+	aufs_read_lock(dentry, AuLock_DW);
+	parent = dentry->d_parent; /* dir inode is locked */
+	di_write_lock_parent(parent);
+	err = au_test_empty(dentry, whlist);
+	//err = -1;
+	if (unlikely(err))
+		goto out_args;
+
+	bstart = au_dbstart(dentry);
+	bwh = au_dbwh(dentry);
+	bindex = -1;
+	wh_dentry = lock_hdir_create_wh(dentry, /*isdir*/1, &bindex, &dt);
+	//wh_dentry = ERR_PTR(-1);
+	err = PTR_ERR(wh_dentry);
+	if (IS_ERR(wh_dentry))
+		goto out_args;
+
+	AuDebugOn(au_dbstart(dentry) != bstart);
+	h_dentry = au_h_dptr(dentry, bstart);
+	dget(h_dentry);
+
+	rmdir_later = 0;
+	if (bindex == bstart) {
+		h_dir = h_dentry->d_parent->d_inode; /* dir inode is locked */
+		IMustLock(h_dir);
+		err = renwh_and_rmdir(dentry, bstart, whlist, dir);
+		//err = -1;
+		if (err > 0) {
+			rmdir_later = err;
+			err = 0;
+		}
+	} else {
+		/* dir inode is locked */
+		AuDebugOn(!wh_dentry
+			  || wh_dentry->d_parent != au_h_dptr(parent, bindex));
+		h_dir = wh_dentry->d_parent->d_inode;
+		IMustLock(h_dir);
+		err = 0;
+	}
+
+	sb = dentry->d_sb;
+	mnt_flags = au_mntflags(sb);
+	if (!err) {
+		//aufs_bindex_t bi, bend;
+
+		if (unlikely(au_opt_test(mnt_flags, UDBA_INOTIFY)
+			     && rmdir_later))
+			au_reset_hinotify(inode, /*flags*/0);
+		clear_nlink(inode);
+		au_set_dbdiropq(dentry, -1);
+		epilog(dir, dentry, bindex);
+
+		if (rmdir_later) {
+			au_whtmp_kick_rmdir(h_dentry, whlist, bstart, dir,
+					    inode, /*noself*/1, args);
+			args = NULL;
+		}
+
+		goto out_unlock; /* success */
+	}
+
+	/* revert */
+	LKTRLabel(revert);
+	if (wh_dentry) {
+		int rerr;
+		unsigned int rev_flags;
+
+		rev_flags = 0;
+		if (unlikely(au_opt_test_dlgt(mnt_flags)))
+			au_fset_rev(rev_flags, DLGT);
+		rerr = do_revert(err, wh_dentry, dentry, bwh, &dt, rev_flags);
+		if (rerr)
+			err = rerr;
+	}
+
+ out_unlock:
+	au_hdir_unlock(h_dir, dir, bindex);
+	dput(wh_dentry);
+	dput(h_dentry);
+ out_args:
+	di_write_unlock(parent);
+	aufs_read_unlock(dentry, AuLock_DW);
+	kfree(args);
+ out_whlist:
+	au_nhash_del(whlist);
+ out:
+	AuTraceErr(err);
+	return err;
+}
-- 
1.4.4.4

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

Powered by blists - more mailing lists