/*
- * Copyright (c) 1999 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 1999-2007 Apple Inc. All rights reserved.
*
- * @APPLE_LICENSE_HEADER_START@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License"). You may not use this file except in compliance with the
- * License. Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.
*
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ *
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT. Please see the
- * License for the specific language governing rights and limitations
- * under the License.
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and
+ * limitations under the License.
*
- * @APPLE_LICENSE_HEADER_END@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
*/
/*
* Copyright (c) 1989, 1993
*
*
* hfs_lookup.c -- code to handle directory traversal on HFS/HFS+ volume
- *
- * MODIFICATION HISTORY:
- * 21-May-1999 Don Brady Add support for HFS rooting.
- * 25-Feb-1999 Clark Warner Fixed the error case of VFS_VGGET when
- * processing DotDot (..) to relock parent
- * 23-Feb-1999 Pat Dirks Finish cleanup around Don's last fix in "." and ".." handling.
- * 11-Nov-1998 Don Brady Take out VFS_VGET that got added as part of previous fix.
- * 14-Oct-1998 Don Brady Fix locking policy volation in hfs_lookup for ".." case
- * (radar #2279902).
- * 4-Jun-1998 Pat Dirks Split off from hfs_vnodeops.c
*/
#include <sys/param.h>
-#include <sys/namei.h>
-#include <sys/buf.h>
#include <sys/file.h>
#include <sys/mount.h>
#include <sys/vnode.h>
#include <sys/malloc.h>
-#include <sys/paths.h>
-
-#include "hfs.h"
-#include "hfs_dbg.h"
-#include "hfscommon/headers/FileMgrInternal.h"
-
-u_int16_t GetForkFromName(struct componentname *cnp);
-int hfs_vget_sibling(struct vnode *vdp, u_int16_t forkType, struct vnode **vpp);
-int hfs_vget_catinfo(struct vnode *parent_vp, struct hfsCatalogInfo *catInfo, u_int32_t forkType, struct vnode **target_vpp);
-
-/*
- * XXX SER fork strings.
- * Put these someplace better
- */
-#define gHFSForkIdentStr "/"
-#define gDataForkNameStr "data"
-#define gRsrcForkNameStr "rsrc"
-
-
-#if DBG_VOP_TEST_LOCKS
-extern void DbgVopTest(int maxSlots, int retval, VopDbgStoreRec *VopDbgStore, char *funcname);
-#endif
+#include <sys/kdebug.h>
+#include <sys/kauth.h>
+#include <sys/namei.h>
-/*****************************************************************************
-*
-* Operations on vnodes
-*
-*****************************************************************************/
+#include "hfs.h"
+#include "hfs_catalog.h"
+#include "hfs_cnode.h"
/*
* FROM FREEBSD 3.1
- * Convert a component of a pathname into a pointer to a locked hfsnode.
+ * Convert a component of a pathname into a pointer to a locked cnode.
* This is a very central and rather complicated routine.
* If the file system is not maintained in a strict tree hierarchy,
* this can result in a deadlock situation (see comments in code below).
* creating, renaming, or deleting a directory entry may be calculated.
* Notice that these are the only operations that can affect the directory of the target.
*
- * If flag has LOCKPARENT or'ed into it and the target of the pathname
- * exists, lookup returns both the target and its parent directory locked.
- * When creating or renaming and LOCKPARENT is specified, the target may
- * not be ".". When deleting and LOCKPARENT is specified, the target may
- * be "."., but the caller must check to ensure it does an vrele and vput
- * instead of two vputs.
- *
* LOCKPARENT and WANTPARENT actually refer to the parent of the last item,
* so if ISLASTCN is not set, they should be ignored. Also they are mutually exclusive, or
* WANTPARENT really implies DONTLOCKPARENT. Either of them set means that the calling
* routine wants to access the parent of the target, locked or unlocked.
*
* Keeping the parent locked as long as possible protects from other processes
- * looking up the same item, so it has to be locked until the hfsnode is totally finished
- *
- * This routine is actually used as VOP_CACHEDLOOKUP method, and the
- * filesystem employs the generic hfs_cache_lookup() as VOP_LOOKUP
- * method.
+ * looking up the same item, so it has to be locked until the cnode is totally finished
*
* hfs_cache_lookup() performs the following for us:
* check that it is a directory
* drop it
* else
* return name.
- * return VOP_CACHEDLOOKUP()
+ * return hfs_lookup()
*
* Overall outline of hfs_lookup:
*
* found:
* if at end of path and deleting, return information to allow delete
* if at end of path and rewriting (RENAME and LOCKPARENT), lock target
- * inode and return info to allow rewrite
+ * cnode and return info to allow rewrite
* if not at end, add name to cache; if at end and neither creating
* nor deleting, add name to cache
*/
+
/*
- * Lookup *nm in directory *pvp, return it in *a_vpp.
- * **a_vpp is held on exit.
- * We create a hfsnode for the file, but we do NOT open the file here.
+ * Lookup *cnp in directory *dvp, return it in *vpp.
+ * **vpp is held on exit.
+ * We create a cnode for the file, but we do NOT open the file here.
#% lookup dvp L ? ?
#% lookup vpp - L -
IN struct vnode *dvp - Parent node of file;
- INOUT struct vnode **vpp - node of target file, its a new node if the target vnode did not exist;
+ INOUT struct vnode **vpp - node of target file, its a new node if
+ the target vnode did not exist;
IN struct componentname *cnp - Name of file;
* When should we lock parent_hp in here ??
*/
-
-int
-hfs_lookup(ap)
- struct vop_cachedlookup_args /* {
- struct vnode *a_dvp;
- struct vnode **a_vpp;
- struct componentname *a_cnp;
- } */ *ap;
+static int
+hfs_lookup(struct vnode *dvp, struct vnode **vpp, struct componentname *cnp, int *cnode_locked)
{
- struct vnode *parent_vp;
- struct vnode *target_vp;
- struct vnode *tparent_vp;
- struct hfsnode *parent_hp; /* parent */
- struct componentname *cnp;
- struct ucred *cred;
- struct proc *p;
- struct hfsCatalogInfo catInfo;
- u_int32_t parent_id;
- u_int32_t nodeID;
- u_int16_t targetLen;
- u_int16_t forkType;
- int flags;
- int lockparent; /* !0 => lockparent flag is set */
- int wantparent; /* !0 => wantparent or lockparent flag */
- int nameiop;
- int retval;
- u_char isDot, isDotDot, found;
- DBG_FUNC_NAME("lookup");
- DBG_VOP_LOCKS_DECL(2);
- DBG_VOP_LOCKS_INIT(0,ap->a_dvp, VOPDBG_LOCKED, VOPDBG_IGNORE, VOPDBG_IGNORE, VOPDBG_POS);
- DBG_VOP_LOCKS_INIT(1,*ap->a_vpp, VOPDBG_IGNORE, VOPDBG_LOCKED, VOPDBG_IGNORE, VOPDBG_POS);
- DBG_VOP_PRINT_FUNCNAME();DBG_VOP_CONT(("\n"));
- DBG_HFS_NODE_CHECK(ap->a_dvp);
-
-
- /*
- * Do initial setup
- */
- INIT_CATALOGDATA(&catInfo.nodeData, 0);
- parent_vp = ap->a_dvp;
- cnp = ap->a_cnp;
- parent_hp = VTOH(parent_vp); /* parent */
- target_vp = NULL;
- targetLen = cnp->cn_namelen;
- nameiop = cnp->cn_nameiop;
- cred = cnp->cn_cred;
- p = cnp->cn_proc;
- lockparent = cnp->cn_flags & LOCKPARENT;
- wantparent = cnp->cn_flags & (LOCKPARENT|WANTPARENT);
- flags = cnp->cn_flags;
- parent_id = H_FILEID(parent_hp);
- nodeID = kUnknownID;
- found = FALSE;
- isDot = FALSE;
- isDotDot = FALSE;
- retval = E_NONE;
- forkType = kUndefinedFork;
-
-
- /*
- * We now have a segment name to search for, and a directory to search.
- *
- */
+ struct cnode *dcp; /* cnode for directory being searched */
+ struct vnode *tvp; /* target vnode */
+ struct hfsmount *hfsmp;
+ int flags;
+ int nameiop;
+ int retval = 0;
+ int isDot;
+ struct cat_desc desc;
+ struct cat_desc cndesc;
+ struct cat_attr attr;
+ struct cat_fork fork;
+ int lockflags;
+
+ retry:
+ dcp = NULL;
+ hfsmp = VTOHFS(dvp);
+ *vpp = NULL;
+ *cnode_locked = 0;
+ isDot = FALSE;
+ tvp = NULL;
+ nameiop = cnp->cn_nameiop;
+ flags = cnp->cn_flags;
+ bzero(&desc, sizeof(desc));
/*
* First check to see if it is a . or .., else look it up.
*/
-
- if (flags & ISDOTDOT) { /* Wanting the parent */
- isDotDot = TRUE;
- found = TRUE; /* .. is always defined */
- nodeID = H_DIRID(parent_hp);
- } /* Wanting ourselves */
- else if ((cnp->cn_nameptr[0] == '.') && (targetLen == 1)) {
+ if (flags & ISDOTDOT) { /* Wanting the parent */
+ cnp->cn_flags &= ~MAKEENTRY;
+ goto found; /* .. is always defined */
+ } else if ((cnp->cn_nameptr[0] == '.') && (cnp->cn_namelen == 1)) {
isDot = TRUE;
- found = TRUE; /* We always know who we are */
- }
- else { /* Wanting something else */
- catInfo.hint = kNoHint;
+ cnp->cn_flags &= ~MAKEENTRY;
+ goto found; /* We always know who we are */
+ } else {
+ if (hfs_lock(VTOC(dvp), HFS_EXCLUSIVE_LOCK) != 0) {
+ retval = ENOENT; /* The parent no longer exists ? */
+ goto exit;
+ }
+ dcp = VTOC(dvp);
- /* lock catalog b-tree */
- retval = hfs_metafilelocking(VTOHFS(parent_vp), kHFSCatalogFileID, LK_SHARED, p);
- if (retval)
- goto Err_Exit;
+ if (dcp->c_flag & C_DIR_MODIFICATION) {
+ // XXXdbg - if we could msleep on a lck_rw_t then we would do that
+ // but since we can't we have to unlock, delay for a bit
+ // and then retry...
+ // msleep((caddr_t)&dcp->c_flag, &dcp->c_rwlock, PINOD, "hfs_vnop_lookup", 0);
+ hfs_unlock(dcp);
+ tsleep((caddr_t)dvp, PRIBIO, "hfs_lookup", 1);
- retval = hfs_getcatalog (VTOVCB(parent_vp), parent_id, cnp->cn_nameptr, targetLen, &catInfo);
-
- /* unlock catalog b-tree */
- (void) hfs_metafilelocking(VTOHFS(parent_vp), kHFSCatalogFileID, LK_RELEASE, p);
-
- if (retval == E_NONE)
- found = TRUE;
- };
+ goto retry;
+ }
+ /* No need to go to catalog if there are no children */
+ if (dcp->c_entries == 0) {
+ goto notfound;
+ }
- /*
- * At this point we know IF we have a valid dir/name.
- */
+ bzero(&cndesc, sizeof(cndesc));
+ cndesc.cd_nameptr = (const u_int8_t *)cnp->cn_nameptr;
+ cndesc.cd_namelen = cnp->cn_namelen;
+ cndesc.cd_parentcnid = dcp->c_fileid;
+ cndesc.cd_hint = dcp->c_childhint;
+ lockflags = hfs_systemfile_lock(hfsmp, SFL_CATALOG, HFS_SHARED_LOCK);
- retval = E_NONE;
- if (! found) {
- /*
- * This is a non-existing entry
- *
- * If creating, and at end of pathname and current
- * directory has not been removed, then can consider
- * allowing file to be created.
- */
- if ((nameiop == CREATE || nameiop == RENAME ||
- (nameiop == DELETE &&
- (ap->a_cnp->cn_flags & DOWHITEOUT) &&
- (ap->a_cnp->cn_flags & ISWHITEOUT))) &&
- (flags & ISLASTCN)) {
- /*
- * Access for write is interpreted as allowing
- * creation of files in the directory.
- */
- retval = VOP_ACCESS(parent_vp, VWRITE, cred, cnp->cn_proc);
- if (retval)
- return (retval);
-
- cnp->cn_flags |= SAVENAME;
- if (!lockparent)
- VOP_UNLOCK(parent_vp, 0, p);
- retval = EJUSTRETURN;
- goto Err_Exit;
- }
-
- /*
- * Insert name into cache (as non-existent) if appropriate.
- */
+ retval = cat_lookup(hfsmp, &cndesc, 0, &desc, &attr, &fork, NULL);
+
+ hfs_systemfile_unlock(hfsmp, lockflags);
+ if (retval == 0) {
+ dcp->c_childhint = desc.cd_hint;
+ /*
+ * Note: We must drop the parent lock here before calling
+ * hfs_getnewvnode (which takes the child lock).
+ */
+ hfs_unlock(dcp);
+ dcp = NULL;
+ goto found;
+ }
+notfound:
/*
- * XXX SER - Here we would store the name in cache as non-existant if not trying to create it, but,
- * the name cache IS case-sensitive, thus maybe showing a negative hit, when the name
- * is only different by case. So hfs does not support negative caching. Something to look at.
- * (See radar 2293594 for a failed example)
- if ((cnp->cn_flags & MAKEENTRY) && nameiop != CREATE)
- cache_enter(parent_vp, *vpp, cnp);
- */
-
- retval = ENOENT;
- }
- else {
- /*
- * We have found an entry
+ * ENAMETOOLONG supersedes other errors
*
- * Here we have to decide what type of vnode to create.
- * There are 3 type of objects that are given:
- * 1. '.': return the same dp
- * 2. '..' return the parent of dp, always a VDIR
- * 3. catinfo rec: return depending on type:
- * A. VDIR, nodeType is kCatalogFolderNode
- * B. VLINK nodeType is kCatalogFileNode, the mode is IFLNK (esp. if it is a link to a directory e.g. bar/link/foo)
- * C. VREG, nodeType is kCatalogFileNode, forkType at this point is unknown
- * To determine the forkType, we can use this algorithm (\0 in the strings mean the NULL character):
- * a. forkType is kDataType iff ISLASTCN is set (as in the case of the default fork e.g. data/foo).
- * b. forkType is kDataType iff ISLASTCN is not set and the namePtr is followed by "/?AppleHFSFork/data\0"
- * c. forkType is kRsrcType iff ISLASTCN is not set and the namePtr is followed by "/?AppleHFSFork/rsrc\0"
- * If the latter two are correct, then we 'consume' the remaining of the name buffer
- * and set the cnp as appropriate.
- * Anything else returns an retval
+ * For a CREATE or RENAME operation on the last component
+ * the ENAMETOOLONG will be handled in the next VNOP.
*/
-
-
+ if ((retval != ENAMETOOLONG) &&
+ (cnp->cn_namelen > kHFSPlusMaxFileNameChars) &&
+ (((flags & ISLASTCN) == 0) || ((nameiop != CREATE) && (nameiop != RENAME)))) {
+ retval = ENAMETOOLONG;
+ } else if (retval == 0) {
+ retval = ENOENT;
+ }
+ if (retval != ENOENT)
+ goto exit;
/*
- * If deleting, and at end of pathname, return
- * parameters which can be used to remove file.
- * If the wantparent flag isn't set, we return only
- * the directory (in ndp->ndvp), otherwise we go
- * on and lock the hfsnode, being careful with ".".
+ * This is a non-existing entry
*
- * Forks cannot be deleted so scan-ahead is illegal, so just return the default fork
+ * If creating, and at end of pathname and current
+ * directory has not been removed, then can consider
+ * allowing file to be created.
*/
- if (nameiop == DELETE && (flags & ISLASTCN)) {
- /*
- * Write access to directory required to delete files.
- */
- retval = VOP_ACCESS(parent_vp, VWRITE, cred, cnp->cn_proc);
- if (retval)
- goto Err_Exit;
-
- if (isDot) { /* Want to return ourselves */
- VREF(parent_vp);
- target_vp = parent_vp;
- goto Err_Exit;
- }
- else if (isDotDot) {
- retval = VFS_VGET(parent_vp->v_mount, &nodeID, &target_vp);
- if (retval)
- goto Err_Exit;
- }
- else {
- retval = hfs_vget_catinfo(parent_vp, &catInfo, kAnyFork, &target_vp);
- if (retval)
- goto Err_Exit;
- CLEAN_CATALOGDATA(&catInfo.nodeData);
- };
-
-
- /*
- * If directory is "sticky", then user must own
- * the directory, or the file in it, else she
- * may not delete it (unless she's root). This
- * implements append-only directories.
- */
- if ((parent_hp->h_meta->h_mode & ISVTX) &&
- (cred->cr_uid != 0) &&
- (cred->cr_uid != parent_hp->h_meta->h_uid) &&
- (target_vp->v_type != VLNK) &&
- (hfs_owner_rights(target_vp, cred, p, false))) {
- vput(target_vp);
- retval = EPERM;
- goto Err_Exit;
- }
-#if HFS_HARDLINKS
- /*
- * If this is a link node then we need to save the name
- * (of the link) so we can delete it from the catalog b-tree.
- * In this case, hfs_remove will then free the component name.
- */
- if (target_vp && (VTOH(target_vp)->h_meta->h_metaflags & IN_DATANODE))
- cnp->cn_flags |= SAVENAME;
-#endif
-
- if (!lockparent)
- VOP_UNLOCK(parent_vp, 0, p);
- goto Err_Exit;
- };
-
+ if ((nameiop == CREATE || nameiop == RENAME ||
+ (nameiop == DELETE &&
+ (cnp->cn_flags & DOWHITEOUT) &&
+ (cnp->cn_flags & ISWHITEOUT))) &&
+ (flags & ISLASTCN) &&
+ !(ISSET(dcp->c_flag, C_DELETED | C_NOEXISTS))) {
+ retval = EJUSTRETURN;
+ goto exit;
+ }
/*
- * If rewriting 'RENAME', return the hfsnode and the
- * information required to rewrite the present directory
+ * Insert name into the name cache (as non-existent).
*/
- if (nameiop == RENAME && wantparent && (cnp->cn_flags & ISLASTCN)) {
-
- if ((retval = VOP_ACCESS(parent_vp, VWRITE, cred, cnp->cn_proc)) != 0)
- goto Err_Exit;
+ if ((hfsmp->hfs_flags & HFS_STANDARD) == 0 &&
+ (cnp->cn_flags & MAKEENTRY) &&
+ (nameiop != CREATE)) {
+ cache_enter(dvp, NULL, cnp);
+ dcp->c_flag |= C_NEG_ENTRIES;
+ }
+ goto exit;
+ }
- /*
- * Careful about locking second inode.
- * This can only occur if the target is ".". like 'mv foo/bar foo/.'
- */
+found:
+ if (flags & ISLASTCN) {
+ switch(nameiop) {
+ case DELETE:
+ cnp->cn_flags &= ~MAKEENTRY;
+ break;
+
+ case RENAME:
+ cnp->cn_flags &= ~MAKEENTRY;
if (isDot) {
retval = EISDIR;
- goto Err_Exit;
+ goto exit;
}
- else if (isDotDot) {
- retval = VFS_VGET(parent_vp->v_mount, &nodeID, &target_vp);
- if (retval)
- goto Err_Exit;
- }
- else {
- /* If then name differs in case, then act like it does not exist
- * This allows renaming foo->Foo
- * Exclude length difference due to compose/decompe issues.
- */
- if ((cnp->cn_namelen == catInfo.nodeData.cnm_length) &&
- strncmp(cnp->cn_nameptr, catInfo.nodeData.cnm_nameptr, targetLen)) {
- if (!lockparent)
- VOP_UNLOCK(parent_vp, 0, p);
- retval = EJUSTRETURN;
- goto Err_Exit;
- };
-
- retval = hfs_vget_catinfo(parent_vp, &catInfo, kAnyFork, &target_vp);
- if (retval)
- goto Err_Exit;
-
- CLEAN_CATALOGDATA(&catInfo.nodeData); /* Should do nothing */
- };
-
- cnp->cn_flags |= SAVENAME;
- if (!lockparent)
- VOP_UNLOCK(parent_vp, 0, p);
-
- goto Err_Exit;
- /* Finished...all is well, goto the end */
- };
-
+ break;
+ }
+ }
+
+ if (isDot) {
+ if ((retval = vnode_get(dvp)))
+ goto exit;
+ *vpp = dvp;
+ } else if (flags & ISDOTDOT) {
/*
- * Step through the translation in the name. We do not `vput' the
- * directory because we may need it again if a symbolic link
- * is relative to the current directory. Instead we save it
- * unlocked as "tparent_vp". We must get the target hfsnode before unlocking
- * the directory to insure that the hfsnode will not be removed
- * before we get it. We prevent deadlock by always fetching
- * inodes from the root, moving down the directory tree. Thus
- * when following backward pointers ".." we must unlock the
- * parent directory before getting the requested directory.
- * There is a potential race condition here if both the current
- * and parent directories are removed before the VFS_VGET for the
- * hfsnode associated with ".." returns. We hope that this occurs
- * infrequently since we cannot avoid this race condition without
- * implementing a sophisticated deadlock detection algorithm.
- * Note also that this simple deadlock detection scheme will not
- * work if the file system has any hard links other than ".."
- * that point backwards in the directory structure.
+ * Directory hard links can have multiple parents so
+ * find the appropriate parent for the current thread.
*/
-
- tparent_vp = parent_vp;
- if (isDotDot) {
- VOP_UNLOCK(tparent_vp, 0, p); /* race to get the inode */
- if ((retval = VFS_VGET(parent_vp->v_mount, &nodeID, &target_vp))) {
- vn_lock(tparent_vp, LK_EXCLUSIVE | LK_RETRY, p);
- goto Err_Exit;
+ if ((retval = hfs_vget(hfsmp, hfs_currentparent(VTOC(dvp)), &tvp, 0))) {
+ goto exit;
}
- if (lockparent && (flags & ISLASTCN) && (tparent_vp != target_vp) &&
- (retval = vn_lock(tparent_vp, LK_EXCLUSIVE, p))) {
- vput(target_vp);
- goto Err_Exit;
- }
+ *cnode_locked = 1;
+ *vpp = tvp;
+ } else {
+ int type = (attr.ca_mode & S_IFMT);
+#if NAMEDRSRCFORK
+ int rsrc_warn = 0;
+
+ /*
+ * Check if caller wants the resource fork but utilized
+ * the legacy "file/rsrc" access path.
+ *
+ * This is deprecated behavior and support for it will not
+ * be allowed beyond case insensitive HFS+ and even that
+ * support will be removed in the next major OS release.
+ */
+ if ((type == S_IFREG) &&
+ ((flags & ISLASTCN) == 0) &&
+ (cnp->cn_nameptr[cnp->cn_namelen] == '/') &&
+ (bcmp(&cnp->cn_nameptr[cnp->cn_namelen+1], "rsrc", 5) == 0) &&
+ ((hfsmp->hfs_flags & (HFS_STANDARD | HFS_CASE_SENSITIVE)) == 0)) {
+
+ cnp->cn_consume = 5;
+ cnp->cn_flags |= CN_WANTSRSRCFORK | ISLASTCN | NOCACHE;
+ cnp->cn_flags &= ~MAKEENTRY;
+ flags |= ISLASTCN;
+ rsrc_warn = 1;
}
- else if (isDot) {
- VREF(parent_vp); /* we want ourself, ie "." */
- target_vp = parent_vp;
+#endif
+ if (!(flags & ISLASTCN) && (type != S_IFDIR) && (type != S_IFLNK)) {
+ retval = ENOTDIR;
+ goto exit;
}
- else {
- mode_t mode;
- /*
- * Determine what fork to get, currenty 3 scenarios are supported:
- * 1. ./foo: if it is a dir, return a VDIR else return data fork
- * 2. ./foo/.__Fork/data: return data fork
- * 3. ./foo/.__Fork/rsrc: return resource fork
- * So the algorithm is:
- * If the object is a directory
- * then return a VDIR vnode
- * else if ISLASTCN is true
- * then get the vnode with forkType=kDataFork
- * else
- * compare with the remaining cnp buffer with "/.__Fork/"
- * if a match
- * then compare string after that with either 'data' or 'rsrc'
- * if match
- * then
- * 'consume' rest of cnp, setting appropriate values and flags
- * return vnode depending on match
- * else
- * bad fork name
- * else
- * illegal path after a file object
- */
-
- mode = (mode_t)(catInfo.nodeData.cnd_mode);
-
- if (catInfo.nodeData.cnd_type == kCatalogFolderNode) {
- forkType = kDirectory; /* Really ignored */
- }
- else if ((mode & IFMT) == IFLNK) {
- forkType = kDataFork;
- } /* After this point, nodeType should be a file */
- else if (flags & ISLASTCN) { /* Create a default fork */
- forkType = kDataFork;
+ /* Don't cache directory hardlink names. */
+ if (attr.ca_recflags & kHFSHasLinkChainMask) {
+ cnp->cn_flags &= ~MAKEENTRY;
}
- else { /* determine what fork was specified */
- forkType = GetForkFromName(cnp);
- flags |= ISLASTCN; /* To know to unlock the parent if needed */
- }; /* else */
-
-
- /* If couldn't determine what type of fork, leave */
- if (forkType == kUndefinedFork) {
- retval = EISDIR;
- goto Err_Exit;
- };
-
- /* Get the vnode now that what type of fork is known */
- DBG_ASSERT((forkType==kDirectory) || (forkType==kDataFork) || (forkType==kRsrcFork));
- retval = hfs_vget_catinfo(tparent_vp, &catInfo, forkType, &target_vp);
- if (retval != E_NONE)
- goto Err_Exit;
-
- if (!lockparent || !(flags & ISLASTCN))
- VOP_UNLOCK(tparent_vp, 0, p);
-
- CLEAN_CATALOGDATA(&catInfo.nodeData);
-
- }; /* else found */
+ /* Names with composed chars are not cached. */
+ if (cnp->cn_namelen != desc.cd_namelen)
+ cnp->cn_flags &= ~MAKEENTRY;
+ retval = hfs_getnewvnode(hfsmp, dvp, cnp, &desc, 0, &attr, &fork, &tvp);
- /*
- * Insert name in cache if wanted.
- * Names with composed chars are not put into the name cache.
- */
- if ((cnp->cn_flags & MAKEENTRY)
- && (cnp->cn_namelen == catInfo.nodeData.cnm_length)) {
+ if (retval) {
/*
- * XXX SER - Might be good idea to bcopy(catInfo.nodeData.fsspec.name, cnp->cn_nameptr)
- * to "normalize" the name cache. This will avoid polluting the name cache with
- * names that are different in case, and allow negative caching
+ * If this was a create operation lookup and another
+ * process removed the object before we had a chance
+ * to create the vnode, then just treat it as the not
+ * found case above and return EJUSTRETURN.
*/
- cache_enter(parent_vp, target_vp, cnp);
+ if ((retval == ENOENT) &&
+ (cnp->cn_nameiop == CREATE) &&
+ (flags & ISLASTCN)) {
+ retval = EJUSTRETURN;
}
-
-
-
- }; /* else found == TRUE */
-
-Err_Exit:
-
- CLEAN_CATALOGDATA(&catInfo.nodeData); /* Just to make sure */
- *ap->a_vpp = target_vp;
-
- DBG_VOP_UPDATE_VP(1, *ap->a_vpp);
- //DBG_VOP_LOOKUP_TEST (funcname, cnp, parent_vp, target_vp);
- //DBG_VOP_LOCKS_TEST(E_NONE);
+ goto exit;
+ }
+ /* Save the origin info of a directory link for future ".." requests. */
+ if (S_ISDIR(attr.ca_mode) && (attr.ca_recflags & kHFSHasLinkChainMask)) {
+ hfs_savelinkorigin(VTOC(tvp), VTOC(dvp)->c_fileid);
+ }
+ *cnode_locked = 1;
+ *vpp = tvp;
+#if NAMEDRSRCFORK
+ if (rsrc_warn) {
+ if ((VTOC(tvp)->c_flag & C_WARNED_RSRC) == 0) {
+ VTOC(tvp)->c_flag |= C_WARNED_RSRC;
+ printf("%.200s: file access by '/rsrc' was deprecated in 10.4\n",
+ cnp->cn_nameptr);
+ }
+ }
+#endif
+ }
+exit:
+ if (dcp) {
+ hfs_unlock(dcp);
+ }
+ cat_releasedesc(&desc);
return (retval);
}
/*
- * Based on vn_cache_lookup (which is vfs_cache_lookup in FreeBSD 3.1)
- *
* Name caching works as follows:
*
* Names found by directory scans are retained in a cache
* is for DELETE, or NOCACHE is set (rewrite), and the
* name is located in the cache, it will be dropped.
*
- * In hfs, since a name can represent multiple forks, it cannot
- * be known what fork the name matches, so further checks have to be done.
- * Currently a policy of first requested, is the one stored, is followed.
- *
- * SER XXX If this proves inadequate maybe we can munge the name to contain a fork reference
- * like foo -> foo.d for the data fork.
*/
+#define S_IXALL 0000111
+
+__private_extern__
int
-hfs_cache_lookup(ap)
- struct vop_lookup_args /* {
- struct vnode *a_dvp;
- struct vnode **a_vpp;
- struct componentname *a_cnp;
- } */ *ap;
+hfs_vnop_lookup(struct vnop_lookup_args *ap)
{
- struct vnode *vdp;
- struct vnode *pdp;
- int lockparent;
+ struct vnode *dvp = ap->a_dvp;
+ struct vnode *vp;
+ struct cnode *cp;
+ struct cnode *dcp;
int error;
struct vnode **vpp = ap->a_vpp;
struct componentname *cnp = ap->a_cnp;
- struct ucred *cred = cnp->cn_cred;
int flags = cnp->cn_flags;
- struct proc *p = cnp->cn_proc;
- struct hfsnode *hp;
- u_int32_t vpid; /* capability number of vnode */
- DBG_FUNC_NAME("cache_lookup");
- DBG_VOP_LOCKS_DECL(2);
- DBG_VOP_LOCKS_INIT(0,ap->a_dvp, VOPDBG_LOCKED, VOPDBG_IGNORE, VOPDBG_IGNORE, VOPDBG_POS);
- DBG_VOP_LOCKS_INIT(1,*ap->a_vpp, VOPDBG_IGNORE, VOPDBG_LOCKED, VOPDBG_IGNORE, VOPDBG_POS);
- DBG_VOP_PRINT_FUNCNAME();DBG_VOP_CONT(("\n"));
- DBG_VOP_CONT(("\tTarget: "));DBG_VOP_PRINT_CPN_INFO(ap->a_cnp);DBG_VOP_CONT(("\n"));
- DBG_HFS_NODE_CHECK(ap->a_dvp);
+ int cnode_locked;
*vpp = NULL;
- vdp = ap->a_dvp;
- lockparent = flags & LOCKPARENT;
-
- if (vdp->v_type != VDIR)
- return (ENOTDIR);
-
- if ((flags & ISLASTCN) && (vdp->v_mount->mnt_flag & MNT_RDONLY) &&
- (cnp->cn_nameiop == DELETE || cnp->cn_nameiop == RENAME))
- return (EROFS);
-
- error = VOP_ACCESS(vdp, VEXEC, cred, cnp->cn_proc);
-
- if (error)
- return (error);
+ dcp = VTOC(dvp);
/*
* Lookup an entry in the cache
- * If the lookup succeeds, the vnode is returned in *vpp, and a status of -1 is
- * returned. If the lookup determines that the name does not exist
- * (negative cacheing), a status of ENOENT is returned. If the lookup
- * fails, a status of zero is returned.
+ *
+ * If the lookup succeeds, the vnode is returned in *vpp,
+ * and a status of -1 is returned.
+ *
+ * If the lookup determines that the name does not exist
+ * (negative cacheing), a status of ENOENT is returned.
+ *
+ * If the lookup fails, a status of zero is returned.
*/
- error = cache_lookup(vdp, vpp, cnp);
-
- if (error == 0) { /* Unsuccessfull */
- DBG_VOP(("\tWas not in name cache\n"));
- error = hfs_lookup(ap);
-#if HFS_HARDLINKS
- if (error)
- return (error);
- /*
- * If this is a hard-link vnode then we need to update
- * the name (of the link) and update the parent ID. This
- * enables getattrlist calls to return correct link info.
- */
- hp = VTOH(*ap->a_vpp);
- if ((flags & ISLASTCN) && (hp->h_meta->h_metaflags & IN_DATANODE)) {
- H_DIRID(hp) = H_FILEID(VTOH(ap->a_dvp));
- hfs_set_metaname(cnp->cn_nameptr, hp->h_meta, HTOHFS(hp));
- }
-#endif
- return (error);
- };
-
- DBG_VOP(("\tName was found in the name cache"));
- if (error == ENOENT) {
- DBG_VOP_CONT((" though it was a NEGATIVE HIT\n"));
- return (error);
- };
- DBG_VOP_CONT(("\n"));
-
-#if HFS_HARDLINKS
+ error = cache_lookup(dvp, vpp, cnp);
+ if (error != -1) {
+ if ((error == ENOENT) && (cnp->cn_nameiop != CREATE))
+ goto exit; /* found a negative cache entry */
+ goto lookup; /* did not find it in the cache */
+ }
/*
- * If this is a hard-link vnode then we need to update
- * the name (of the link) and update the parent ID. This
- * enables getattrlist calls to return correct link info.
+ * We have a name that matched
+ * cache_lookup returns the vp with an iocount reference already taken
*/
- hp = VTOH(*vpp);
- if ((flags & ISLASTCN) && (hp->h_meta->h_metaflags & IN_DATANODE)) {
- H_DIRID(hp) = H_FILEID(VTOH(vdp));
- hfs_set_metaname(cnp->cn_nameptr, hp->h_meta, HTOHFS(hp));
- }
-#endif
+ error = 0;
+ vp = *vpp;
- /* We have a name that matched */
- pdp = vdp;
- vdp = *vpp;
- vpid = vdp->v_id;
- if (pdp == vdp) { /* lookup on "." */
- VREF(vdp);
- error = 0;
- } else if (flags & ISDOTDOT) {
- /*
- * Carefull on the locking policy,
- * remember we always lock from parent to child, so have
- * to release lock on child before trying to lock parent
- * then regain lock if needed
- */
- VOP_UNLOCK(pdp, 0, p);
- error = vget(vdp, LK_EXCLUSIVE, p);
- if (!error && lockparent && (flags & ISLASTCN))
- error = vn_lock(pdp, LK_EXCLUSIVE, p);
- } else {
- /*
- * Check to see if a specific fork is not being requested.
- *
- * If it is a file and not the last path item
- * then check if its a proper fork
- * If it is, check to see if the matched vnode is the same fork
- * else see if the proper fork exists.
- * If it does, return that one, else do VOP_CACHEDLOOKUP()
- * Notice that nothing is done if an undefined fork is named. Just leave and let lookup()
- * handle strange cases.
- *
- * XXX SER Notice that when the target is not what was in the name cache,
- * it is locked, before trying to get its sibling. Could this be a problem since both
- * siblings can be locked, but not in a determinalistic order????
- */
- u_int16_t forkType;
-
- error = vget(vdp, LK_EXCLUSIVE, p);
- if ((! error) && (vdp->v_type == VREG) && (vpid == vdp->v_id)) {
- if (!(flags & ISLASTCN)) {
- forkType = GetForkFromName(cnp);
- if (forkType != kUndefinedFork) {
- flags |= ISLASTCN;
- if (H_FORKTYPE(VTOH(vdp)) != forkType) {
- error = hfs_vget_sibling(vdp, forkType, vpp);
- vput(vdp);
- if (! error) {
- vdp = *vpp;
- vpid = vdp->v_id;
- }
- }
- }
- }
- else {
- /* Its the last item, so we want the data fork */
- if (H_FORKTYPE(VTOH(vdp)) != kDataFork) {
- error = hfs_vget_sibling(vdp, kDataFork, vpp);
- vput(vdp);
- if (! error) {
- vdp = *vpp;
- vpid = vdp->v_id;
- }
- }
- };
- };
- if (!lockparent || error || !(flags & ISLASTCN))
- VOP_UNLOCK(pdp, 0, p);
- };
/*
- * Check that the capability number did not change
- * while we were waiting for the lock.
+ * If this is a hard-link vnode then we need to update
+ * the name (of the link), the parent ID, the cnid, the
+ * text encoding and the catalog hint. This enables
+ * getattrlist calls to return the correct link info.
*/
- if (!error) {
- if (vpid == vdp->v_id)
- return (0); /* HERE IS THE NORMAL EXIT FOR CACHE LOOKUP!!!! */
- /*
- * The above is the NORMAL exit, after this point is an error
- * condition.
- */
- vput(vdp);
- if (lockparent && pdp != vdp && (flags & ISLASTCN))
- VOP_UNLOCK(pdp, 0, p);
- }
- error = vn_lock(pdp, LK_EXCLUSIVE, p);
- if (error)
- return (error);
- return (hfs_lookup(ap));
-}
-
-/*
- * Parses a componentname and sees if the remaining path
- * contains a hfs named fork specifier. If it does set the
- * componentname to consume the rest of the path, and
- * return the forkType
- */
-
-u_int16_t GetForkFromName(struct componentname *cnp)
-{
- u_int16_t forkType = kUndefinedFork;
- char *tcp = cnp->cn_nameptr + cnp->cn_namelen;
-
- if (bcmp(tcp, _PATH_FORKSPECIFIER, sizeof(_PATH_FORKSPECIFIER) - 1) == 0) {
- /* Its a HFS fork, so far */
- tcp += (sizeof(_PATH_FORKSPECIFIER) - 1);
- if (bcmp(tcp, _PATH_DATANAME, sizeof(_PATH_DATANAME)) == 0) {
- forkType = kDataFork;
- cnp->cn_consume = sizeof(_PATH_FORKSPECIFIER) + sizeof(_PATH_DATANAME) - 2;
- }
- else if (bcmp(tcp, _PATH_RSRCNAME, sizeof(_PATH_RSRCNAME)) == 0) {
- forkType = kRsrcFork;
- cnp->cn_consume = sizeof(_PATH_FORKSPECIFIER) + sizeof(_PATH_RSRCNAME) - 2;
- }; /* else if */
- }; /* if bcmp */
-
-
- /* XXX SER For backwards compatability...keep it */
- if (forkType == kUndefinedFork) {
- tcp = cnp->cn_nameptr + cnp->cn_namelen;
- if (bcmp(tcp, gHFSForkIdentStr, sizeof(gHFSForkIdentStr) - 1) == 0) {
- /* Its a HFS fork, so far */
- tcp += (sizeof(gHFSForkIdentStr) - 1);
- if (bcmp(tcp, gDataForkNameStr, sizeof(gDataForkNameStr)) == 0) {
- forkType = kDataFork;
- cnp->cn_consume = sizeof(gHFSForkIdentStr) + sizeof(gDataForkNameStr) - 2;
- }
- else if (bcmp(tcp, gRsrcForkNameStr, sizeof(gRsrcForkNameStr)) == 0) {
- forkType = kRsrcFork;
- cnp->cn_consume = sizeof(gHFSForkIdentStr) + sizeof(gRsrcForkNameStr) - 2;
- }; /* else if */
- }; /* if bcmp */
- };
-
- return forkType;
-}
+ cp = VTOC(vp);
-#if DBG_VOP_TEST_LOCKS
+ if ((flags & ISLASTCN) && (cp->c_flag & C_HARDLINK)) {
+ hfs_lock(cp, HFS_FORCE_LOCK);
+ if ((cp->c_parentcnid != dcp->c_cnid) ||
+ (bcmp(cnp->cn_nameptr, cp->c_desc.cd_nameptr, cp->c_desc.cd_namelen) != 0)) {
+ struct cat_desc desc;
+ int lockflags;
-void DbgLookupTest( char *funcname, struct componentname *cnp, struct vnode *dvp, struct vnode *vp)
-{
- if (! (hfs_dbg_lookup || hfs_dbg_all))
- return;
-
-
- if (dvp) {
- if (lockstatus(&VTOH(dvp)->h_lock)) {
- DBG_LOOKUP (("%s: Parent vnode exited LOCKED", funcname));
- }
- else {
- DBG_LOOKUP (("%s: Parent vnode exited UNLOCKED", funcname));
+ /*
+ * Get an updated descriptor
+ */
+ desc.cd_nameptr = (const u_int8_t *)cnp->cn_nameptr;
+ desc.cd_namelen = cnp->cn_namelen;
+ desc.cd_parentcnid = dcp->c_fileid;
+ desc.cd_hint = dcp->c_childhint;
+ desc.cd_encoding = 0;
+ desc.cd_cnid = 0;
+ desc.cd_flags = S_ISDIR(cp->c_mode) ? CD_ISDIR : 0;
+
+ lockflags = hfs_systemfile_lock(VTOHFS(dvp), SFL_CATALOG, HFS_SHARED_LOCK);
+ if (cat_lookup(VTOHFS(vp), &desc, 0, &desc, NULL, NULL, NULL) == 0)
+ replace_desc(cp, &desc);
+ hfs_systemfile_unlock(VTOHFS(dvp), lockflags);
}
+ hfs_unlock(cp);
}
-
- if (vp) {
- if (vp==dvp)
- {
- DBG_LOOKUP (("%s: Target and Parent are the same", funcname));
- }
- else {
- if (lockstatus(&VTOH(vp)->h_lock)) {
- DBG_LOOKUP (("%s: Found vnode exited LOCKED", funcname));
- }
- else {
- DBG_LOOKUP (("%s: Found vnode exited LOCKED", funcname));
- }
+#if NAMEDRSRCFORK
+ /*
+ * Check if caller wants the resource fork but utilized
+ * the legacy "file/rsrc" access path.
+ *
+ * This is deprecated behavior and support for it will not
+ * be allowed beyond case insensitive HFS+ and even that
+ * support will be removed in the next major OS release.
+ */
+ if ((dvp != vp) &&
+ ((flags & ISLASTCN) == 0) &&
+ vnode_isreg(vp) &&
+ (cnp->cn_nameptr[cnp->cn_namelen] == '/') &&
+ (bcmp(&cnp->cn_nameptr[cnp->cn_namelen+1], "rsrc", 5) == 0) &&
+ ((VTOHFS(vp)->hfs_flags & (HFS_STANDARD | HFS_CASE_SENSITIVE)) == 0)) {
+ cnp->cn_consume = 5;
+ cnp->cn_flags |= CN_WANTSRSRCFORK | ISLASTCN | NOCACHE;
+ cnp->cn_flags &= ~MAKEENTRY;
+
+ hfs_lock(cp, HFS_FORCE_LOCK);
+ if ((cp->c_flag & C_WARNED_RSRC) == 0) {
+ cp->c_flag |= C_WARNED_RSRC;
+ printf("%.200s: file access by '/rsrc' was deprecated in 10.4\n", cnp->cn_nameptr);
}
- DBG_LOOKUP (("%s: Found vnode 0x%x has vtype of %d\n ", funcname, (u_int)vp, vp->v_type));
+ hfs_unlock(cp);
}
- else
- DBG_LOOKUP (("%s: Found vnode exited NULL\n", funcname));
-
+#endif
+ return (error);
+
+lookup:
+ /*
+ * The vnode was not in the name cache or it was stale.
+ *
+ * So we need to do a real lookup.
+ */
+ cnode_locked = 0;
+ error = hfs_lookup(dvp, vpp, cnp, &cnode_locked);
+
+ if (cnode_locked)
+ hfs_unlock(VTOC(*vpp));
+exit:
+ return (error);
}
-#endif /* DBG_VOP_TEST_LOCKS */