X-Git-Url: https://git.saurik.com/apple/xnu.git/blobdiff_plain/d7e50217d7adf6e52786a38bcaa4cd698cb9a79e..c18c124eaa464aaaa5549e99e5a70fc9cbb50944:/bsd/vfs/vfs_vnops.c diff --git a/bsd/vfs/vfs_vnops.c b/bsd/vfs/vfs_vnops.c index 40177b401..9b431080f 100644 --- a/bsd/vfs/vfs_vnops.c +++ b/bsd/vfs/vfs_vnops.c @@ -1,16 +1,19 @@ /* - * Copyright (c) 2000-2002 Apple Computer, Inc. All rights reserved. + * Copyright (c) 2000-2014 Apple Inc. All rights reserved. * - * @APPLE_LICENSE_HEADER_START@ - * - * Copyright (c) 1999-2003 Apple Computer, Inc. All Rights Reserved. + * @APPLE_OSREFERENCE_LICENSE_HEADER_START@ * * 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. Please obtain a copy of the License at - * http://www.opensource.apple.com/apsl/ and read it before using this - * file. + * 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. + * + * 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 @@ -20,7 +23,7 @@ * 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) 1995 NeXT Computer, Inc. All Rights Reserved */ /* @@ -63,333 +66,935 @@ * @(#)vfs_vnops.c 8.14 (Berkeley) 6/15/95 * */ +/* + * NOTICE: This file was modified by SPARTA, Inc. in 2005 to introduce + * support for mandatory and extensible security protections. This notice + * is included in support of clause 2.2 (b) of the Apple Public License, + * Version 2.0. + */ #include +#include #include #include -#include +#include #include -#include -#include -#include +#include +#include +#include #include -#include +#include #include #include +/* Temporary workaround for ubc.h until +#undef ubc_setcred +int ubc_setcred(struct vnode *, struct proc *); #include #include +#include +#include +#include +#include +#include +#include +#include #include +#include #include +#include -static int vn_closefile __P((struct file *fp, struct proc *p)); -static int vn_ioctl __P((struct file *fp, u_long com, caddr_t data, - struct proc *p)); -static int vn_read __P((struct file *fp, struct uio *uio, - struct ucred *cred, int flags, struct proc *p)); -static int vn_write __P((struct file *fp, struct uio *uio, - struct ucred *cred, int flags, struct proc *p)); -static int vn_select __P(( struct file *fp, int which, void * wql, - struct proc *p)); +#if CONFIG_MACF +#include +#endif + +#if CONFIG_PROTECT +#include +#endif + +extern void sigpup_attach_vnode(vnode_t); /* XXX */ + +static int vn_closefile(struct fileglob *fp, vfs_context_t ctx); +static int vn_ioctl(struct fileproc *fp, u_long com, caddr_t data, + vfs_context_t ctx); +static int vn_read(struct fileproc *fp, struct uio *uio, int flags, + vfs_context_t ctx); +static int vn_write(struct fileproc *fp, struct uio *uio, int flags, + vfs_context_t ctx); +static int vn_select( struct fileproc *fp, int which, void * wql, + vfs_context_t ctx); +static int vn_kqfilt_add(struct fileproc *fp, struct knote *kn, + vfs_context_t ctx); +static void filt_vndetach(struct knote *kn); +static int filt_vnode(struct knote *kn, long hint); +static int vn_open_auth_finish(vnode_t vp, int fmode, vfs_context_t ctx); +#if 0 +static int vn_kqfilt_remove(struct vnode *vp, uintptr_t ident, + vfs_context_t ctx); +#endif -struct fileops vnops = - { vn_read, vn_write, vn_ioctl, vn_select, vn_closefile }; +const struct fileops vnops = { + DTYPE_VNODE, + vn_read, + vn_write, + vn_ioctl, + vn_select, + vn_closefile, + vn_kqfilt_add, + NULL +}; + +struct filterops vnode_filtops = { + .f_isfd = 1, + .f_attach = NULL, + .f_detach = filt_vndetach, + .f_event = filt_vnode +}; /* * Common code for vnode open operations. - * Check permissions, and call the VOP_OPEN or VOP_CREATE routine. + * Check permissions, and call the VNOP_OPEN or VNOP_CREATE routine. + * + * XXX the profusion of interfaces here is probably a bad thing. + */ +int +vn_open(struct nameidata *ndp, int fmode, int cmode) +{ + return(vn_open_modflags(ndp, &fmode, cmode)); +} + +int +vn_open_modflags(struct nameidata *ndp, int *fmodep, int cmode) +{ + struct vnode_attr va; + + VATTR_INIT(&va); + VATTR_SET(&va, va_mode, cmode); + + return(vn_open_auth(ndp, fmodep, &va)); +} + +static int +vn_open_auth_finish(vnode_t vp, int fmode, vfs_context_t ctx) +{ + int error; + + if ((error = vnode_ref_ext(vp, fmode, 0)) != 0) { + goto bad; + } + + /* Call out to allow 3rd party notification of open. + * Ignore result of kauth_authorize_fileop call. + */ +#if CONFIG_MACF + mac_vnode_notify_open(ctx, vp, fmode); +#endif + kauth_authorize_fileop(vfs_context_ucred(ctx), KAUTH_FILEOP_OPEN, + (uintptr_t)vp, 0); + + sigpup_attach_vnode(vp); + + return 0; + +bad: + return error; + +} + +/* + * May do nameidone() to allow safely adding an FSEvent. Cue off of ni_dvp to + * determine whether that has happened. + */ +static int +vn_open_auth_do_create(struct nameidata *ndp, struct vnode_attr *vap, int fmode, boolean_t *did_create, boolean_t *did_open, vfs_context_t ctx) +{ + uint32_t status = 0; + vnode_t dvp = ndp->ni_dvp; + int batched; + int error; + vnode_t vp; + + batched = vnode_compound_open_available(ndp->ni_dvp); + *did_open = FALSE; + + VATTR_SET(vap, va_type, VREG); + if (fmode & O_EXCL) + vap->va_vaflags |= VA_EXCLUSIVE; + +#if NAMEDRSRCFORK + if (ndp->ni_cnd.cn_flags & CN_WANTSRSRCFORK) { + if ((error = vn_authorize_create(dvp, &ndp->ni_cnd, vap, ctx, NULL)) != 0) + goto out; + if ((error = vnode_makenamedstream(dvp, &ndp->ni_vp, XATTR_RESOURCEFORK_NAME, 0, ctx)) != 0) + goto out; + *did_create = TRUE; + } else { +#endif + if (!batched) { + if ((error = vn_authorize_create(dvp, &ndp->ni_cnd, vap, ctx, NULL)) != 0) + goto out; + } + + error = vn_create(dvp, &ndp->ni_vp, ndp, vap, VN_CREATE_DOOPEN, fmode, &status, ctx); + if (error != 0) { + if (batched) { + *did_create = (status & COMPOUND_OPEN_STATUS_DID_CREATE) ? TRUE : FALSE; + } else { + *did_create = FALSE; + } + + if (error == EKEEPLOOKING) { + if (*did_create) { + panic("EKEEPLOOKING, but we did a create?"); + } + if (!batched) { + panic("EKEEPLOOKING from filesystem that doesn't support compound vnops?"); + } + if ((ndp->ni_flag & NAMEI_CONTLOOKUP) == 0) { + panic("EKEEPLOOKING, but continue flag not set?"); + } + + /* + * Do NOT drop the dvp: we need everything to continue the lookup. + */ + return error; + } + } else { + if (batched) { + *did_create = (status & COMPOUND_OPEN_STATUS_DID_CREATE) ? 1 : 0; + *did_open = TRUE; + } else { + *did_create = TRUE; + } + } +#if NAMEDRSRCFORK + } +#endif + + vp = ndp->ni_vp; + + if (*did_create) { + int update_flags = 0; + + // Make sure the name & parent pointers are hooked up + if (vp->v_name == NULL) + update_flags |= VNODE_UPDATE_NAME; + if (vp->v_parent == NULLVP) + update_flags |= VNODE_UPDATE_PARENT; + + if (update_flags) + vnode_update_identity(vp, dvp, ndp->ni_cnd.cn_nameptr, ndp->ni_cnd.cn_namelen, ndp->ni_cnd.cn_hash, update_flags); + + vnode_put(dvp); + ndp->ni_dvp = NULLVP; + +#if CONFIG_FSE + if (need_fsevent(FSE_CREATE_FILE, vp)) { + add_fsevent(FSE_CREATE_FILE, ctx, + FSE_ARG_VNODE, vp, + FSE_ARG_DONE); + } +#endif + } +out: + if (ndp->ni_dvp != NULLVP) { + vnode_put(dvp); + ndp->ni_dvp = NULLVP; + } + + return error; +} + +/* + * Open a file with authorization, updating the contents of the structures + * pointed to by ndp, fmodep, and vap as necessary to perform the requested + * operation. This function is used for both opens of existing files, and + * creation of new files. + * + * Parameters: ndp The nami data pointer describing the + * file + * fmodep A pointer to an int containg the mode + * information to be used for the open + * vap A pointer to the vnode attribute + * descriptor to be used for the open + * + * Indirect: * Contents of the data structures pointed + * to by the parameters are modified as + * necessary to the requested operation. + * + * Returns: 0 Success + * !0 errno value + * + * Notes: The kauth_filesec_t in 'vap', if any, is in host byte order. + * + * The contents of '*ndp' will be modified, based on the other + * arguments to this function, and to return file and directory + * data necessary to satisfy the requested operation. + * + * If the file does not exist and we are creating it, then the + * O_TRUNC flag will be cleared in '*fmodep' to indicate to the + * caller that the file was not truncated. + * + * If the file exists and the O_EXCL flag was not specified, then + * the O_CREAT flag will be cleared in '*fmodep' to indicate to + * the caller that the existing file was merely opened rather + * than created. + * + * The contents of '*vap' will be modified as necessary to + * complete the operation, including setting of supported + * attribute, clearing of fields containing unsupported attributes + * in the request, if the request proceeds without them, etc.. + * + * XXX: This function is too complicated in actings on its arguments + * + * XXX: We should enummerate the possible errno values here, and where + * in the code they originated. */ int -vn_open(ndp, fmode, cmode) - register struct nameidata *ndp; - int fmode, cmode; +vn_open_auth(struct nameidata *ndp, int *fmodep, struct vnode_attr *vap) { - register struct vnode *vp; - register struct proc *p = ndp->ni_cnd.cn_proc; - register struct ucred *cred = p->p_ucred; - struct vattr vat; - struct vattr *vap = &vat; + struct vnode *vp; + struct vnode *dvp; + vfs_context_t ctx = ndp->ni_cnd.cn_context; int error; - int didhold = 0; + int fmode; + uint32_t origcnflags; + boolean_t did_create; + boolean_t did_open; + boolean_t need_vnop_open; + boolean_t batched; + boolean_t ref_failed; + +again: + vp = NULL; + dvp = NULL; + batched = FALSE; + did_create = FALSE; + need_vnop_open = TRUE; + ref_failed = FALSE; + fmode = *fmodep; + origcnflags = ndp->ni_cnd.cn_flags; + /* + * O_CREAT + */ if (fmode & O_CREAT) { + if ( (fmode & O_DIRECTORY) ) { + error = EINVAL; + goto out; + } ndp->ni_cnd.cn_nameiop = CREATE; - ndp->ni_cnd.cn_flags = LOCKPARENT | LOCKLEAF; - if ((fmode & O_EXCL) == 0) +#if CONFIG_TRIGGERS + ndp->ni_op = OP_LINK; +#endif + /* Inherit USEDVP, vnode_open() supported flags only */ + ndp->ni_cnd.cn_flags &= (USEDVP | NOCROSSMOUNT); + ndp->ni_cnd.cn_flags |= LOCKPARENT | LOCKLEAF | AUDITVNPATH1; + ndp->ni_flag = NAMEI_COMPOUNDOPEN; +#if NAMEDRSRCFORK + /* open calls are allowed for resource forks. */ + ndp->ni_cnd.cn_flags |= CN_ALLOWRSRCFORK; +#endif + if ((fmode & O_EXCL) == 0 && (fmode & O_NOFOLLOW) == 0 && (origcnflags & FOLLOW) != 0) ndp->ni_cnd.cn_flags |= FOLLOW; - bwillwrite(); - if (error = namei(ndp)) - return (error); - if (ndp->ni_vp == NULL) { - VATTR_NULL(vap); - vap->va_type = VREG; - vap->va_mode = cmode; - if (fmode & O_EXCL) - vap->va_vaflags |= VA_EXCLUSIVE; - VOP_LEASE(ndp->ni_dvp, p, cred, LEASE_WRITE); - if (error = VOP_CREATE(ndp->ni_dvp, &ndp->ni_vp, - &ndp->ni_cnd, vap)) - return (error); - fmode &= ~O_TRUNC; - vp = ndp->ni_vp; - } else { - VOP_ABORTOP(ndp->ni_dvp, &ndp->ni_cnd); - if (ndp->ni_dvp == ndp->ni_vp) - vrele(ndp->ni_dvp); - else - vput(ndp->ni_dvp); - ndp->ni_dvp = NULL; + +continue_create_lookup: + if ( (error = namei(ndp)) ) + goto out; + + dvp = ndp->ni_dvp; + vp = ndp->ni_vp; + + batched = vnode_compound_open_available(dvp); + + /* not found, create */ + if (vp == NULL) { + /* must have attributes for a new file */ + if (vap == NULL) { + vnode_put(dvp); + error = EINVAL; + goto out; + } + /* + * Attempt a create. For a system supporting compound VNOPs, we may + * find an existing file or create one; in either case, we will already + * have the file open and no VNOP_OPEN() will be needed. + */ + error = vn_open_auth_do_create(ndp, vap, fmode, &did_create, &did_open, ctx); + + dvp = ndp->ni_dvp; vp = ndp->ni_vp; - if (fmode & O_EXCL) { + + /* + * Detected a node that the filesystem couldn't handle. Don't call + * nameidone() yet, because we need that path buffer. + */ + if (error == EKEEPLOOKING) { + if (!batched) { + panic("EKEEPLOOKING from a filesystem that doesn't support compound VNOPs?"); + } + goto continue_create_lookup; + } + + nameidone(ndp); + if (dvp) { + panic("Shouldn't have a dvp here."); + } + + if (error) { + /* + * Check for a creation or unlink race. + */ + if (((error == EEXIST) && !(fmode & O_EXCL)) || + ((error == ENOENT) && (fmode & O_CREAT))){ + if (vp) + vnode_put(vp); + goto again; + } + goto bad; + } + + need_vnop_open = !did_open; + } + else { + if (fmode & O_EXCL) error = EEXIST; + + /* + * We have a vnode. Use compound open if available + * or else fall through to "traditional" path. Note: can't + * do a compound open for root, because the parent belongs + * to a different FS. + */ + if (error == 0 && batched && (vnode_mount(dvp) == vnode_mount(vp))) { + error = VNOP_COMPOUND_OPEN(dvp, &ndp->ni_vp, ndp, 0, fmode, NULL, NULL, ctx); + + if (error == 0) { + vp = ndp->ni_vp; + need_vnop_open = FALSE; + } else if (error == EKEEPLOOKING) { + if ((ndp->ni_flag & NAMEI_CONTLOOKUP) == 0) { + panic("EKEEPLOOKING, but continue flag not set?"); + } + goto continue_create_lookup; + } + } + nameidone(ndp); + vnode_put(dvp); + ndp->ni_dvp = NULLVP; + + if (error) { goto bad; } + fmode &= ~O_CREAT; + + /* Fall through */ } - } else { - ndp->ni_cnd.cn_nameiop = LOOKUP; - ndp->ni_cnd.cn_flags = FOLLOW | LOCKLEAF; - if (error = namei(ndp)) - return (error); - vp = ndp->ni_vp; - } - if (vp->v_type == VSOCK) { - error = EOPNOTSUPP; - goto bad; } + else { + /* + * Not O_CREAT + */ + ndp->ni_cnd.cn_nameiop = LOOKUP; + /* Inherit USEDVP, vnode_open() supported flags only */ + ndp->ni_cnd.cn_flags &= (USEDVP | NOCROSSMOUNT); + ndp->ni_cnd.cn_flags |= FOLLOW | LOCKLEAF | AUDITVNPATH1 | WANTPARENT; +#if NAMEDRSRCFORK + /* open calls are allowed for resource forks. */ + ndp->ni_cnd.cn_flags |= CN_ALLOWRSRCFORK; +#endif + ndp->ni_flag = NAMEI_COMPOUNDOPEN; -#if DIAGNOSTIC - if (UBCINFOMISSING(vp)) - panic("vn_open: ubc_info_init"); -#endif /* DIAGNOSTIC */ + /* preserve NOFOLLOW from vnode_open() */ + if (fmode & O_NOFOLLOW || fmode & O_SYMLINK || (origcnflags & FOLLOW) == 0) { + ndp->ni_cnd.cn_flags &= ~FOLLOW; + } - if (UBCINFOEXISTS(vp) && ((didhold = ubc_hold(vp)) == 0)) { - error = ENOENT; - goto bad; + /* Do a lookup, possibly going directly to filesystem for compound operation */ + do { + if ( (error = namei(ndp)) ) + goto out; + vp = ndp->ni_vp; + dvp = ndp->ni_dvp; + + /* Check for batched lookup-open */ + batched = vnode_compound_open_available(dvp); + if (batched && ((vp == NULLVP) || (vnode_mount(dvp) == vnode_mount(vp)))) { + error = VNOP_COMPOUND_OPEN(dvp, &ndp->ni_vp, ndp, 0, fmode, NULL, NULL, ctx); + vp = ndp->ni_vp; + if (error == 0) { + need_vnop_open = FALSE; + } else if (error == EKEEPLOOKING) { + if ((ndp->ni_flag & NAMEI_CONTLOOKUP) == 0) { + panic("EKEEPLOOKING, but continue flag not set?"); + } + } + } + } while (error == EKEEPLOOKING); + + nameidone(ndp); + vnode_put(dvp); + ndp->ni_dvp = NULLVP; + + if (error) { + goto bad; + } } - if ((fmode & O_CREAT) == 0) { - if (fmode & FREAD && fmode & (FWRITE | O_TRUNC)) { - int err = 0; - if (vp->v_type == VDIR) - err = EISDIR; - else - err = vn_writechk(vp); - if (err && !(error = VOP_ACCESS(vp, VREAD, cred, p))) - error = err; - if (error || (error = VOP_ACCESS(vp, VREAD|VWRITE, - cred, p))) - goto bad; - } else if (fmode & FREAD) { - if ((error = VOP_ACCESS(vp, VREAD, cred, p))) - goto bad; - } else if (fmode & (FWRITE | O_TRUNC)) { - if (vp->v_type == VDIR) { - error = EISDIR; + /* + * By this point, nameidone() is called, dvp iocount is dropped, + * and dvp pointer is cleared. + */ + if (ndp->ni_dvp != NULLVP) { + panic("Haven't cleaned up adequately in vn_open_auth()"); + } + + /* + * Expect to use this code for filesystems without compound VNOPs, for the root + * of a filesystem, which can't be "looked up" in the sense of VNOP_LOOKUP(), + * and for shadow files, which do not live on the same filesystems as their "parents." + */ + if (need_vnop_open) { + if (batched && !vnode_isvroot(vp) && !vnode_isnamedstream(vp)) { + panic("Why am I trying to use VNOP_OPEN() on anything other than the root or a named stream?"); + } + + if (!did_create) { + error = vn_authorize_open_existing(vp, &ndp->ni_cnd, fmode, ctx, NULL); + if (error) { goto bad; } - if ((error = vn_writechk(vp)) || - (error = VOP_ACCESS(vp, VWRITE, cred, p))) + } + +#if CONFIG_PROTECT + /* + * Perform any content protection access checks prior to calling + * into the filesystem, if the raw encrypted mode was not + * requested. + * + * If the va_dataprotect_flags are NOT active, or if they are, + * but they do not have the VA_DP_RAWENCRYPTED bit set, then we need + * to perform the checks. + */ + if (!(VATTR_IS_ACTIVE (vap, va_dataprotect_flags)) || + ((vap->va_dataprotect_flags & VA_DP_RAWENCRYPTED) == 0)) { + error = cp_handle_open (vp, fmode); + if (error) { goto bad; + } } - } - if (fmode & O_TRUNC) { - VOP_UNLOCK(vp, 0, p); /* XXX */ - VOP_LEASE(vp, p, cred, LEASE_WRITE); - (void)vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p); /* XXX */ - VATTR_NULL(vap); - vap->va_size = 0; - if (error = VOP_SETATTR(vp, vap, cred, p)) +#endif + + error = VNOP_OPEN(vp, fmode, ctx); + if (error) { goto bad; + } + need_vnop_open = FALSE; + } + + // if the vnode is tagged VOPENEVT and the current process + // has the P_CHECKOPENEVT flag set, then we or in the O_EVTONLY + // flag to the open mode so that this open won't count against + // the vnode when carbon delete() does a vnode_isinuse() to see + // if a file is currently in use. this allows spotlight + // importers to not interfere with carbon apps that depend on + // the no-delete-if-busy semantics of carbon delete(). + // + if (!did_create && (vp->v_flag & VOPENEVT) && (current_proc()->p_flag & P_CHECKOPENEVT)) { + fmode |= O_EVTONLY; } - if (error = VOP_OPEN(vp, fmode, cred, p)) { + /* + * Grab reference, etc. + */ + error = vn_open_auth_finish(vp, fmode, ctx); + if (error) { + ref_failed = TRUE; goto bad; } - if (fmode & FWRITE) - if (++vp->v_writecount <= 0) - panic("vn_open: v_writecount"); + /* Compound VNOP open is responsible for doing the truncate */ + if (batched || did_create) + fmode &= ~O_TRUNC; + + *fmodep = fmode; return (0); + bad: - VOP_UNLOCK(vp, 0, p); - if (didhold) - ubc_rele(vp); - vrele(vp); + /* Opened either explicitly or by a batched create */ + if (!need_vnop_open) { + VNOP_CLOSE(vp, fmode, ctx); + } + + ndp->ni_vp = NULL; + if (vp) { +#if NAMEDRSRCFORK + /* Aggressively recycle shadow files if we error'd out during open() */ + if ((vnode_isnamedstream(vp)) && + (vp->v_parent != NULLVP) && + (vnode_isshadow(vp))) { + vnode_recycle(vp); + } +#endif + vnode_put(vp); + /* + * Check for a race against unlink. We had a vnode + * but according to vnode_authorize or VNOP_OPEN it + * no longer exists. + * + * EREDRIVEOPEN: means that we were hit by the tty allocation race. + */ + if (((error == ENOENT) && (*fmodep & O_CREAT)) || (error == EREDRIVEOPEN) || ref_failed) { + goto again; + } + } + +out: return (error); } +#if vn_access_DEPRECATED /* - * Check for write permissions on the specified vnode. - * Prototype text segments cannot be written. + * Authorize an action against a vnode. This has been the canonical way to + * ensure that the credential/process/etc. referenced by a vfs_context + * is granted the rights called out in 'mode' against the vnode 'vp'. + * + * Unfortunately, the use of VREAD/VWRITE/VEXEC makes it very difficult + * to add support for more rights. As such, this interface will be deprecated + * and callers will use vnode_authorize instead. */ int -vn_writechk(vp) - register struct vnode *vp; +vn_access(vnode_t vp, int mode, vfs_context_t context) { - - /* - * If there's shared text associated with - * the vnode, try to free it up once. If - * we fail, we can't allow writing. - */ -#if 0 - /* XXXXX Not sure we need this */ - if (vp->v_flag & VTEXT) - return (ETXTBSY); -#endif /* 0 */ - return (0); + kauth_action_t action; + + action = 0; + if (mode & VREAD) + action |= KAUTH_VNODE_READ_DATA; + if (mode & VWRITE) + action |= KAUTH_VNODE_WRITE_DATA; + if (mode & VEXEC) + action |= KAUTH_VNODE_EXECUTE; + + return(vnode_authorize(vp, NULL, action, context)); } +#endif /* vn_access_DEPRECATED */ /* * Vnode close call */ int -vn_close(vp, flags, cred, p) - register struct vnode *vp; - int flags; - struct ucred *cred; - struct proc *p; +vn_close(struct vnode *vp, int flags, vfs_context_t ctx) { int error; + int flusherror = 0; + +#if NAMEDRSRCFORK + /* Sync data from resource fork shadow file if needed. */ + if ((vp->v_flag & VISNAMEDSTREAM) && + (vp->v_parent != NULLVP) && + vnode_isshadow(vp)) { + if (flags & FWASWRITTEN) { + flusherror = vnode_flushnamedstream(vp->v_parent, vp, ctx); + } + } +#endif + + /* work around for foxhound */ + if (vnode_isspec(vp)) + (void)vnode_rele_ext(vp, flags, 0); + + /* + * On HFS, we flush when the last writer closes. We do this + * because resource fork vnodes hold a reference on data fork + * vnodes and that will prevent them from getting VNOP_INACTIVE + * which will delay when we flush cached data. In future, we + * might find it beneficial to do this for all file systems. + * Note that it's OK to access v_writecount without the lock + * in this context. + */ + if (vp->v_tag == VT_HFS && (flags & FWRITE) && vp->v_writecount == 1) + VNOP_FSYNC(vp, MNT_NOWAIT, ctx); + + error = VNOP_CLOSE(vp, flags, ctx); - if (flags & FWRITE) - vp->v_writecount--; - error = VOP_CLOSE(vp, flags, cred, p); - ubc_rele(vp); - vrele(vp); +#if CONFIG_FSE + if (flags & FWASWRITTEN) { + if (need_fsevent(FSE_CONTENT_MODIFIED, vp)) { + add_fsevent(FSE_CONTENT_MODIFIED, ctx, + FSE_ARG_VNODE, vp, + FSE_ARG_DONE); + } + } +#endif + + if (!vnode_isspec(vp)) + (void)vnode_rele_ext(vp, flags, 0); + + if (flusherror) { + error = flusherror; + } return (error); } +static int +vn_read_swapfile( + struct vnode *vp, + uio_t uio) +{ + int error; + off_t swap_count, this_count; + off_t file_end, read_end; + off_t prev_resid; + char *my_swap_page; + + /* + * Reading from a swap file will get you zeroes. + */ + + my_swap_page = NULL; + error = 0; + swap_count = uio_resid(uio); + + file_end = ubc_getsize(vp); + read_end = uio->uio_offset + uio_resid(uio); + if (uio->uio_offset >= file_end) { + /* uio starts after end of file: nothing to read */ + swap_count = 0; + } else if (read_end > file_end) { + /* uio extends beyond end of file: stop before that */ + swap_count -= (read_end - file_end); + } + + while (swap_count > 0) { + if (my_swap_page == NULL) { + MALLOC(my_swap_page, char *, PAGE_SIZE, + M_TEMP, M_WAITOK); + memset(my_swap_page, '\0', PAGE_SIZE); + /* add an end-of-line to keep line counters happy */ + my_swap_page[PAGE_SIZE-1] = '\n'; + } + this_count = swap_count; + if (this_count > PAGE_SIZE) { + this_count = PAGE_SIZE; + } + + prev_resid = uio_resid(uio); + error = uiomove((caddr_t) my_swap_page, + this_count, + uio); + if (error) { + break; + } + swap_count -= (prev_resid - uio_resid(uio)); + } + if (my_swap_page != NULL) { + FREE(my_swap_page, M_TEMP); + my_swap_page = NULL; + } + + return error; +} /* * Package up an I/O request on a vnode into a uio and do it. */ int -vn_rdwr(rw, vp, base, len, offset, segflg, ioflg, cred, aresid, p) - enum uio_rw rw; - struct vnode *vp; - caddr_t base; - int len; - off_t offset; - enum uio_seg segflg; - int ioflg; - struct ucred *cred; - int *aresid; - struct proc *p; +vn_rdwr( + enum uio_rw rw, + struct vnode *vp, + caddr_t base, + int len, + off_t offset, + enum uio_seg segflg, + int ioflg, + kauth_cred_t cred, + int *aresid, + proc_t p) +{ + int64_t resid; + int result; + + result = vn_rdwr_64(rw, + vp, + (uint64_t)(uintptr_t)base, + (int64_t)len, + offset, + segflg, + ioflg, + cred, + &resid, + p); + + /* "resid" should be bounded above by "len," which is an int */ + if (aresid != NULL) { + *aresid = resid; + } + + return result; +} + + +int +vn_rdwr_64( + enum uio_rw rw, + struct vnode *vp, + uint64_t base, + int64_t len, + off_t offset, + enum uio_seg segflg, + int ioflg, + kauth_cred_t cred, + int64_t *aresid, + proc_t p) { - struct uio auio; - struct iovec aiov; + uio_t auio; + int spacetype; + struct vfs_context context; int error=0; + char uio_buf[ UIO_SIZEOF(1) ]; + + context.vc_thread = current_thread(); + context.vc_ucred = cred; + + if (UIO_SEG_IS_USER_SPACE(segflg)) { + spacetype = proc_is64bit(p) ? UIO_USERSPACE64 : UIO_USERSPACE32; + } + else { + spacetype = UIO_SYSSPACE; + } + auio = uio_createwithbuffer(1, offset, spacetype, rw, + &uio_buf[0], sizeof(uio_buf)); + uio_addiov(auio, base, len); - /* FIXME XXX */ - if ((ioflg & IO_NODELOCKED) == 0) - (void)vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p); - auio.uio_iov = &aiov; - auio.uio_iovcnt = 1; - aiov.iov_base = base; - aiov.iov_len = len; - auio.uio_resid = len; - auio.uio_offset = offset; - auio.uio_segflg = segflg; - auio.uio_rw = rw; - auio.uio_procp = p; - - if (rw == UIO_READ) - error = VOP_READ(vp, &auio, ioflg, cred); - else - error = VOP_WRITE(vp, &auio, ioflg, cred); +#if CONFIG_MACF + /* XXXMAC + * IO_NOAUTH should be re-examined. + * Likely that mediation should be performed in caller. + */ + if ((ioflg & IO_NOAUTH) == 0) { + /* passed cred is fp->f_cred */ + if (rw == UIO_READ) + error = mac_vnode_check_read(&context, cred, vp); + else + error = mac_vnode_check_write(&context, cred, vp); + } +#endif + + if (error == 0) { + if (rw == UIO_READ) { + if (vnode_isswap(vp) && ((ioflg & IO_SWAP_DISPATCH) == 0)) { + error = vn_read_swapfile(vp, auio); + } else { + error = VNOP_READ(vp, auio, ioflg, &context); + } + } else { + error = VNOP_WRITE(vp, auio, ioflg, &context); + } + } if (aresid) - *aresid = auio.uio_resid; + *aresid = uio_resid(auio); else - if (auio.uio_resid && error == 0) + if (uio_resid(auio) && error == 0) error = EIO; - if ((ioflg & IO_NODELOCKED) == 0) - VOP_UNLOCK(vp, 0, p); return (error); } +static inline void +vn_offset_lock(struct fileglob *fg) +{ + lck_mtx_lock_spin(&fg->fg_lock); + while (fg->fg_lflags & FG_OFF_LOCKED) { + fg->fg_lflags |= FG_OFF_LOCKWANT; + msleep(&fg->fg_lflags, &fg->fg_lock, PVFS | PSPIN, + "fg_offset_lock_wait", 0); + } + fg->fg_lflags |= FG_OFF_LOCKED; + lck_mtx_unlock(&fg->fg_lock); +} + +static inline void +vn_offset_unlock(struct fileglob *fg) +{ + int lock_wanted = 0; + + lck_mtx_lock_spin(&fg->fg_lock); + if (fg->fg_lflags & FG_OFF_LOCKWANT) { + lock_wanted = 1; + } + fg->fg_lflags &= ~(FG_OFF_LOCKED | FG_OFF_LOCKWANT); + lck_mtx_unlock(&fg->fg_lock); + if (lock_wanted) { + wakeup(&fg->fg_lflags); + } +} + /* * File table vnode read routine. */ static int -vn_read(fp, uio, cred, flags, p) - struct file *fp; - struct uio *uio; - struct ucred *cred; - int flags; - struct proc *p; +vn_read(struct fileproc *fp, struct uio *uio, int flags, vfs_context_t ctx) { struct vnode *vp; - int error, ioflag; + int error; + int ioflag; off_t count; + int offset_locked = 0; - if (p != uio->uio_procp) - panic("vn_read: uio_procp does not match p"); + vp = (struct vnode *)fp->f_fglob->fg_data; + if ( (error = vnode_getwithref(vp)) ) { + return(error); + } - vp = (struct vnode *)fp->f_data; - ioflag = 0; - if (fp->f_flag & FNONBLOCK) - ioflag |= IO_NDELAY; - VOP_LEASE(vp, p, cred, LEASE_READ); - error = vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p); - if (error) +#if CONFIG_MACF + error = mac_vnode_check_read(ctx, vfs_context_ucred(ctx), vp); + if (error) { + (void)vnode_put(vp); return (error); - if ((flags & FOF_OFFSET) == 0) - uio->uio_offset = fp->f_offset; - count = uio->uio_resid; - - if(UBCINFOEXISTS(vp)) { - memory_object_t pager; - struct iovec *iov; - off_t file_off; - kern_return_t kr = KERN_SUCCESS; - kern_return_t ret = KERN_SUCCESS; - int count; - - pager = (memory_object_t)ubc_getpager(vp); - file_off = uio->uio_offset; - iov = uio->uio_iov; - count = uio->uio_iovcnt; - while(count) { - kr = vm_conflict_check(current_map(), - (vm_offset_t)iov->iov_base, iov->iov_len, - pager, file_off); - if(kr == KERN_ALREADY_WAITING) { - if((count != uio->uio_iovcnt) && - (ret != KERN_ALREADY_WAITING)) { - error = EINVAL; - goto done; - } - ret = KERN_ALREADY_WAITING; - } else if (kr != KERN_SUCCESS) { - error = EINVAL; - goto done; - } - if(kr != ret) { - error = EINVAL; - goto done; - } - file_off += iov->iov_len; - iov++; - count--; - } - if(ret == KERN_ALREADY_WAITING) { - uio->uio_resid = 0; - if ((flags & FOF_OFFSET) == 0) - fp->f_offset += - count - uio->uio_resid; - error = 0; - goto done; - } - } - error = VOP_READ(vp, uio, ioflag, cred); - if ((flags & FOF_OFFSET) == 0) - fp->f_offset += count - uio->uio_resid; -done: - VOP_UNLOCK(vp, 0, p); + } +#endif + + /* This signals to VNOP handlers that this read came from a file table read */ + ioflag = IO_SYSCALL_DISPATCH; + + if (fp->f_fglob->fg_flag & FNONBLOCK) + ioflag |= IO_NDELAY; + if ((fp->f_fglob->fg_flag & FNOCACHE) || vnode_isnocache(vp)) + ioflag |= IO_NOCACHE; + if (fp->f_fglob->fg_flag & FENCRYPTED) { + ioflag |= IO_ENCRYPTED; + } + if (fp->f_fglob->fg_flag & FNORDAHEAD) + ioflag |= IO_RAOFF; + + if ((flags & FOF_OFFSET) == 0) { + if ((vnode_vtype(vp) == VREG) && !vnode_isswap(vp)) { + vn_offset_lock(fp->f_fglob); + offset_locked = 1; + } + uio->uio_offset = fp->f_fglob->fg_offset; + } + count = uio_resid(uio); + + if (vnode_isswap(vp)) { + /* special case for swap files */ + error = vn_read_swapfile(vp, uio); + } else { + error = VNOP_READ(vp, uio, ioflag, ctx); + } + if ((flags & FOF_OFFSET) == 0) { + fp->f_fglob->fg_offset += count - uio_resid(uio); + if (offset_locked) { + vn_offset_unlock(fp->f_fglob); + offset_locked = 0; + } + } + + (void)vnode_put(vp); return (error); } @@ -398,129 +1003,221 @@ done: * File table vnode write routine. */ static int -vn_write(fp, uio, cred, flags, p) - struct file *fp; - struct uio *uio; - struct ucred *cred; - int flags; - struct proc *p; +vn_write(struct fileproc *fp, struct uio *uio, int flags, vfs_context_t ctx) { struct vnode *vp; int error, ioflag; off_t count; + int clippedsize = 0; + int partialwrite=0; + int residcount, oldcount; + int offset_locked = 0; + proc_t p = vfs_context_proc(ctx); - if (p != uio->uio_procp) - panic("vn_write: uio_procp does not match p"); + count = 0; + vp = (struct vnode *)fp->f_fglob->fg_data; + if ( (error = vnode_getwithref(vp)) ) { + return(error); + } + +#if CONFIG_MACF + error = mac_vnode_check_write(ctx, vfs_context_ucred(ctx), vp); + if (error) { + (void)vnode_put(vp); + return (error); + } +#endif - vp = (struct vnode *)fp->f_data; - ioflag = IO_UNIT; - if (vp->v_type == VREG) - bwillwrite(); - if (vp->v_type == VREG && (fp->f_flag & O_APPEND)) + /* + * IO_SYSCALL_DISPATCH signals to VNOP handlers that this write came from + * a file table write + */ + ioflag = (IO_UNIT | IO_SYSCALL_DISPATCH); + + if (vp->v_type == VREG && (fp->f_fglob->fg_flag & O_APPEND)) ioflag |= IO_APPEND; - if (fp->f_flag & FNONBLOCK) + if (fp->f_fglob->fg_flag & FNONBLOCK) ioflag |= IO_NDELAY; - if ((fp->f_flag & O_FSYNC) || - (vp->v_mount && (vp->v_mount->mnt_flag & MNT_SYNCHRONOUS))) + if ((fp->f_fglob->fg_flag & FNOCACHE) || vnode_isnocache(vp)) + ioflag |= IO_NOCACHE; + if (fp->f_fglob->fg_flag & FNODIRECT) + ioflag |= IO_NODIRECT; + if (fp->f_fglob->fg_flag & FSINGLE_WRITER) + ioflag |= IO_SINGLE_WRITER; + + /* + * Treat synchronous mounts and O_FSYNC on the fd as equivalent. + * + * XXX We treat O_DSYNC as O_FSYNC for now, since we can not delay + * XXX the non-essential metadata without some additional VFS work; + * XXX the intent at this point is to plumb the interface for it. + */ + if ((fp->f_fglob->fg_flag & (O_FSYNC|O_DSYNC)) || + (vp->v_mount && (vp->v_mount->mnt_flag & MNT_SYNCHRONOUS))) { ioflag |= IO_SYNC; - VOP_LEASE(vp, p, cred, LEASE_WRITE); - error = vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, p); - if (error) - return (error); + } + if ((flags & FOF_OFFSET) == 0) { - uio->uio_offset = fp->f_offset; - count = uio->uio_resid; - } - - if(UBCINFOEXISTS(vp)) { - memory_object_t pager; - struct iovec *iov; - off_t file_off; - kern_return_t kr = KERN_SUCCESS; - kern_return_t ret = KERN_SUCCESS; - int count; - - pager = (memory_object_t)ubc_getpager(vp); - file_off = uio->uio_offset; - iov = uio->uio_iov; - count = uio->uio_iovcnt; - while(count) { - kr = vm_conflict_check(current_map(), - (vm_offset_t)iov->iov_base, - iov->iov_len, pager, file_off); - if(kr == KERN_ALREADY_WAITING) { - if((count != uio->uio_iovcnt) && - (ret != KERN_ALREADY_WAITING)) { - error = EINVAL; - goto done; - } - ret = KERN_ALREADY_WAITING; - } else if (kr != KERN_SUCCESS) { - error = EINVAL; - goto done; - } - if(kr != ret) { - error = EINVAL; - goto done; - } - file_off += iov->iov_len; - iov++; - count--; + if ((vnode_vtype(vp) == VREG) && !vnode_isswap(vp)) { + vn_offset_lock(fp->f_fglob); + offset_locked = 1; + } + uio->uio_offset = fp->f_fglob->fg_offset; + count = uio_resid(uio); + } + if (((flags & FOF_OFFSET) == 0) && + vfs_context_proc(ctx) && (vp->v_type == VREG) && + (((rlim_t)(uio->uio_offset + uio_resid(uio)) > p->p_rlimit[RLIMIT_FSIZE].rlim_cur) || + ((rlim_t)uio_resid(uio) > (p->p_rlimit[RLIMIT_FSIZE].rlim_cur - uio->uio_offset)))) { + /* + * If the requested residual would cause us to go past the + * administrative limit, then we need to adjust the residual + * down to cause fewer bytes than requested to be written. If + * we can't do that (e.g. the residual is already 1 byte), + * then we fail the write with EFBIG. + */ + residcount = uio_resid(uio); + if ((rlim_t)(uio->uio_offset + uio_resid(uio)) > p->p_rlimit[RLIMIT_FSIZE].rlim_cur) { + clippedsize = (uio->uio_offset + uio_resid(uio)) - p->p_rlimit[RLIMIT_FSIZE].rlim_cur; + } else if ((rlim_t)uio_resid(uio) > (p->p_rlimit[RLIMIT_FSIZE].rlim_cur - uio->uio_offset)) { + clippedsize = (p->p_rlimit[RLIMIT_FSIZE].rlim_cur - uio->uio_offset); } - if(ret == KERN_ALREADY_WAITING) { - uio->uio_resid = 0; - if ((flags & FOF_OFFSET) == 0) - fp->f_offset += - count - uio->uio_resid; - error = 0; - goto done; + if (clippedsize >= residcount) { + psignal(p, SIGXFSZ); + error = EFBIG; + goto error_out; + } + partialwrite = 1; + uio_setresid(uio, residcount-clippedsize); + } + if ((flags & FOF_OFFSET) != 0) { + /* for pwrite, append should be ignored */ + ioflag &= ~IO_APPEND; + if (p && (vp->v_type == VREG) && + ((rlim_t)uio->uio_offset >= p->p_rlimit[RLIMIT_FSIZE].rlim_cur)) { + psignal(p, SIGXFSZ); + error = EFBIG; + goto error_out; + } + if (p && (vp->v_type == VREG) && + ((rlim_t)(uio->uio_offset + uio_resid(uio)) > p->p_rlimit[RLIMIT_FSIZE].rlim_cur)) { + //Debugger("vn_bwrite:overstepping the bounds"); + residcount = uio_resid(uio); + clippedsize = (uio->uio_offset + uio_resid(uio)) - p->p_rlimit[RLIMIT_FSIZE].rlim_cur; + partialwrite = 1; + uio_setresid(uio, residcount-clippedsize); } } - error = VOP_WRITE(vp, uio, ioflag, cred); + + error = VNOP_WRITE(vp, uio, ioflag, ctx); + + if (partialwrite) { + oldcount = uio_resid(uio); + uio_setresid(uio, oldcount + clippedsize); + } if ((flags & FOF_OFFSET) == 0) { if (ioflag & IO_APPEND) - fp->f_offset = uio->uio_offset; + fp->f_fglob->fg_offset = uio->uio_offset; else - fp->f_offset += count - uio->uio_resid; + fp->f_fglob->fg_offset += count - uio_resid(uio); + if (offset_locked) { + vn_offset_unlock(fp->f_fglob); + offset_locked = 0; + } } /* * Set the credentials on successful writes */ if ((error == 0) && (vp->v_tag == VT_NFS) && (UBCINFOEXISTS(vp))) { - ubc_setcred(vp, p); + /* + * When called from aio subsystem, we only have the proc from + * which to get the credential, at this point, so use that + * instead. This means aio functions are incompatible with + * per-thread credentials (aio operations are proxied). We + * can't easily correct the aio vs. settid race in this case + * anyway, so we disallow it. + */ + if ((flags & FOF_PCRED) == 0) { + ubc_setthreadcred(vp, p, current_thread()); + } else { + ubc_setcred(vp, p); + } } + (void)vnode_put(vp); + return (error); -done: - VOP_UNLOCK(vp, 0, p); +error_out: + if (offset_locked) { + vn_offset_unlock(fp->f_fglob); + } + (void)vnode_put(vp); return (error); } /* * File table vnode stat routine. + * + * Returns: 0 Success + * EBADF + * ENOMEM + * vnode_getattr:??? */ int -vn_stat(vp, sb, p) - struct vnode *vp; - register struct stat *sb; - struct proc *p; +vn_stat_noauth(struct vnode *vp, void *sbptr, kauth_filesec_t *xsec, int isstat64, vfs_context_t ctx) { - struct vattr vattr; - register struct vattr *vap; + struct vnode_attr va; int error; u_short mode; + kauth_filesec_t fsec; + struct stat *sb = (struct stat *)0; /* warning avoidance ; protected by isstat64 */ + struct stat64 * sb64 = (struct stat64 *)0; /* warning avoidance ; protected by isstat64 */ - vap = &vattr; - error = VOP_GETATTR(vp, vap, p->p_ucred, p); + if (isstat64 != 0) + sb64 = (struct stat64 *)sbptr; + else + sb = (struct stat *)sbptr; + memset(&va, 0, sizeof(va)); + VATTR_INIT(&va); + VATTR_WANTED(&va, va_fsid); + VATTR_WANTED(&va, va_fileid); + VATTR_WANTED(&va, va_mode); + VATTR_WANTED(&va, va_type); + VATTR_WANTED(&va, va_nlink); + VATTR_WANTED(&va, va_uid); + VATTR_WANTED(&va, va_gid); + VATTR_WANTED(&va, va_rdev); + VATTR_WANTED(&va, va_data_size); + VATTR_WANTED(&va, va_access_time); + VATTR_WANTED(&va, va_modify_time); + VATTR_WANTED(&va, va_change_time); + VATTR_WANTED(&va, va_create_time); + VATTR_WANTED(&va, va_flags); + VATTR_WANTED(&va, va_gen); + VATTR_WANTED(&va, va_iosize); + /* lower layers will synthesise va_total_alloc from va_data_size if required */ + VATTR_WANTED(&va, va_total_alloc); + if (xsec != NULL) { + VATTR_WANTED(&va, va_uuuid); + VATTR_WANTED(&va, va_guuid); + VATTR_WANTED(&va, va_acl); + } + error = vnode_getattr(vp, &va, ctx); if (error) - return (error); + goto out; /* * Copy from vattr table */ - sb->st_dev = vap->va_fsid; - sb->st_ino = vap->va_fileid; - mode = vap->va_mode; + if (isstat64 != 0) { + sb64->st_dev = va.va_fsid; + sb64->st_ino = (ino64_t)va.va_fileid; + + } else { + sb->st_dev = va.va_fsid; + sb->st_ino = (ino_t)va.va_fileid; + } + mode = va.va_mode; switch (vp->v_type) { case VREG: mode |= S_IFREG; @@ -544,150 +1241,579 @@ vn_stat(vp, sb, p) mode |= S_IFIFO; break; default: - return (EBADF); + error = EBADF; + goto out; }; - sb->st_mode = mode; - sb->st_nlink = vap->va_nlink; - sb->st_uid = vap->va_uid; - sb->st_gid = vap->va_gid; - sb->st_rdev = vap->va_rdev; - sb->st_size = vap->va_size; - sb->st_atimespec = vap->va_atime; - sb->st_mtimespec = vap->va_mtime; - sb->st_ctimespec = vap->va_ctime; - sb->st_blksize = vap->va_blocksize; - sb->st_flags = vap->va_flags; + if (isstat64 != 0) { + sb64->st_mode = mode; + sb64->st_nlink = VATTR_IS_SUPPORTED(&va, va_nlink) ? (u_int16_t)va.va_nlink : 1; + sb64->st_uid = va.va_uid; + sb64->st_gid = va.va_gid; + sb64->st_rdev = va.va_rdev; + sb64->st_size = va.va_data_size; + sb64->st_atimespec = va.va_access_time; + sb64->st_mtimespec = va.va_modify_time; + sb64->st_ctimespec = va.va_change_time; + sb64->st_birthtimespec = + VATTR_IS_SUPPORTED(&va, va_create_time) ? va.va_create_time : va.va_change_time; + sb64->st_blksize = va.va_iosize; + sb64->st_flags = va.va_flags; + sb64->st_blocks = roundup(va.va_total_alloc, 512) / 512; + } else { + sb->st_mode = mode; + sb->st_nlink = VATTR_IS_SUPPORTED(&va, va_nlink) ? (u_int16_t)va.va_nlink : 1; + sb->st_uid = va.va_uid; + sb->st_gid = va.va_gid; + sb->st_rdev = va.va_rdev; + sb->st_size = va.va_data_size; + sb->st_atimespec = va.va_access_time; + sb->st_mtimespec = va.va_modify_time; + sb->st_ctimespec = va.va_change_time; + sb->st_blksize = va.va_iosize; + sb->st_flags = va.va_flags; + sb->st_blocks = roundup(va.va_total_alloc, 512) / 512; + } + + /* if we're interested in extended security data and we got an ACL */ + if (xsec != NULL) { + if (!VATTR_IS_SUPPORTED(&va, va_acl) && + !VATTR_IS_SUPPORTED(&va, va_uuuid) && + !VATTR_IS_SUPPORTED(&va, va_guuid)) { + *xsec = KAUTH_FILESEC_NONE; + } else { + + if (VATTR_IS_SUPPORTED(&va, va_acl) && (va.va_acl != NULL)) { + fsec = kauth_filesec_alloc(va.va_acl->acl_entrycount); + } else { + fsec = kauth_filesec_alloc(0); + } + if (fsec == NULL) { + error = ENOMEM; + goto out; + } + fsec->fsec_magic = KAUTH_FILESEC_MAGIC; + if (VATTR_IS_SUPPORTED(&va, va_uuuid)) { + fsec->fsec_owner = va.va_uuuid; + } else { + fsec->fsec_owner = kauth_null_guid; + } + if (VATTR_IS_SUPPORTED(&va, va_guuid)) { + fsec->fsec_group = va.va_guuid; + } else { + fsec->fsec_group = kauth_null_guid; + } + if (VATTR_IS_SUPPORTED(&va, va_acl) && (va.va_acl != NULL)) { + bcopy(va.va_acl, &(fsec->fsec_acl), KAUTH_ACL_COPYSIZE(va.va_acl)); + } else { + fsec->fsec_acl.acl_entrycount = KAUTH_FILESEC_NOACL; + } + *xsec = fsec; + } + } + /* Do not give the generation number out to unpriviledged users */ - if (suser(p->p_ucred, &p->p_acflag)) - sb->st_gen = 0; - else - sb->st_gen = vap->va_gen; - sb->st_blocks = vap->va_bytes / S_BLKSIZE; - return (0); + if (va.va_gen && !vfs_context_issuser(ctx)) { + if (isstat64 != 0) + sb64->st_gen = 0; + else + sb->st_gen = 0; + } else { + if (isstat64 != 0) + sb64->st_gen = va.va_gen; + else + sb->st_gen = va.va_gen; + } + + error = 0; +out: + if (VATTR_IS_SUPPORTED(&va, va_acl) && va.va_acl != NULL) + kauth_acl_free(va.va_acl); + return (error); } +int +vn_stat(struct vnode *vp, void *sb, kauth_filesec_t *xsec, int isstat64, vfs_context_t ctx) +{ + int error; + +#if CONFIG_MACF + error = mac_vnode_check_stat(ctx, NOCRED, vp); + if (error) + return (error); +#endif + + /* authorize */ + if ((error = vnode_authorize(vp, NULL, KAUTH_VNODE_READ_ATTRIBUTES | KAUTH_VNODE_READ_SECURITY, ctx)) != 0) + return(error); + + /* actual stat */ + return(vn_stat_noauth(vp, sb, xsec, isstat64, ctx)); +} + + /* * File table vnode ioctl routine. */ static int -vn_ioctl(fp, com, data, p) - struct file *fp; - u_long com; - caddr_t data; - struct proc *p; +vn_ioctl(struct fileproc *fp, u_long com, caddr_t data, vfs_context_t ctx) { - register struct vnode *vp = ((struct vnode *)fp->f_data); - struct vattr vattr; + struct vnode *vp = ((struct vnode *)fp->f_fglob->fg_data); + off_t file_size; int error; struct vnode *ttyvp; + struct session * sessp; - switch (vp->v_type) { + if ( (error = vnode_getwithref(vp)) ) { + return(error); + } + +#if CONFIG_MACF + error = mac_vnode_check_ioctl(ctx, vp, com); + if (error) + goto out; +#endif + switch (vp->v_type) { case VREG: case VDIR: if (com == FIONREAD) { - if (error = VOP_GETATTR(vp, &vattr, p->p_ucred, p)) - return (error); - *(int *)data = vattr.va_size - fp->f_offset; - return (0); + if ((error = vnode_size(vp, &file_size, ctx)) != 0) + goto out; + *(int *)data = file_size - fp->f_fglob->fg_offset; + goto out; + } + if (com == FIONBIO || com == FIOASYNC) { /* XXX */ + goto out; } - if (com == FIONBIO || com == FIOASYNC) /* XXX */ - return (0); /* XXX */ /* fall into ... */ default: - return (ENOTTY); + error = ENOTTY; + goto out; case VFIFO: case VCHR: case VBLK: - /* Should not be able to set block size from user space */ - if(com == DKIOCSETBLOCKSIZE) - return (EPERM); - - if (com == FIODTYPE) { - if (vp->v_type == VBLK) { - if (major(vp->v_rdev) >= nblkdev) - return (ENXIO); - *(int *)data = bdevsw[major(vp->v_rdev)].d_type; - } else if (vp->v_type == VCHR) { - if (major(vp->v_rdev) >= nchrdev) - return (ENXIO); - *(int *)data = cdevsw[major(vp->v_rdev)].d_type; - } else { - return (ENOTTY); - } - return (0); - } - error = VOP_IOCTL(vp, com, data, fp->f_flag, p->p_ucred, p); - if (error == 0 && com == TIOCSCTTY) { - VREF(vp); - ttyvp = p->p_session->s_ttyvp; - p->p_session->s_ttyvp = vp; - if (ttyvp) - vrele(ttyvp); - } - return (error); + /* Should not be able to set block size from user space */ + if (com == DKIOCSETBLOCKSIZE) { + error = EPERM; + goto out; + } + + if (com == FIODTYPE) { + if (vp->v_type == VBLK) { + if (major(vp->v_rdev) >= nblkdev) { + error = ENXIO; + goto out; + } + *(int *)data = bdevsw[major(vp->v_rdev)].d_type; + + } else if (vp->v_type == VCHR) { + if (major(vp->v_rdev) >= nchrdev) { + error = ENXIO; + goto out; + } + *(int *)data = cdevsw[major(vp->v_rdev)].d_type; + } else { + error = ENOTTY; + goto out; + } + goto out; + } + error = VNOP_IOCTL(vp, com, data, fp->f_fglob->fg_flag, ctx); + + if (error == 0 && com == TIOCSCTTY) { + sessp = proc_session(vfs_context_proc(ctx)); + + session_lock(sessp); + ttyvp = sessp->s_ttyvp; + sessp->s_ttyvp = vp; + sessp->s_ttyvid = vnode_vid(vp); + session_unlock(sessp); + session_rele(sessp); + } } +out: + (void)vnode_put(vp); + return(error); } /* * File table vnode select routine. */ static int -vn_select(fp, which, wql, p) - struct file *fp; - int which; - void * wql; - struct proc *p; +vn_select(struct fileproc *fp, int which, void *wql, __unused vfs_context_t ctx) { + int error; + struct vnode * vp = (struct vnode *)fp->f_fglob->fg_data; + struct vfs_context context; - return(VOP_SELECT(((struct vnode *)fp->f_data), which, fp->f_flag, - fp->f_cred, wql, p)); + if ( (error = vnode_getwithref(vp)) == 0 ) { + context.vc_thread = current_thread(); + context.vc_ucred = fp->f_fglob->fg_cred; + +#if CONFIG_MACF + /* + * XXX We should use a per thread credential here; minimally, + * XXX the process credential should have a persistent + * XXX reference on it before being passed in here. + */ + error = mac_vnode_check_select(ctx, vp, which); + if (error == 0) +#endif + error = VNOP_SELECT(vp, which, fp->f_fglob->fg_flag, wql, ctx); + + (void)vnode_put(vp); + } + return(error); + } /* - * Check that the vnode is still valid, and if so - * acquire requested lock. + * File table vnode close routine. + */ +static int +vn_closefile(struct fileglob *fg, vfs_context_t ctx) +{ + struct vnode *vp = (struct vnode *)fg->fg_data; + int error; + struct flock lf; + + if ( (error = vnode_getwithref(vp)) == 0 ) { + + if ((fg->fg_flag & FHASLOCK) && + FILEGLOB_DTYPE(fg) == DTYPE_VNODE) { + lf.l_whence = SEEK_SET; + lf.l_start = 0; + lf.l_len = 0; + lf.l_type = F_UNLCK; + + (void)VNOP_ADVLOCK(vp, (caddr_t)fg, F_UNLCK, &lf, F_FLOCK, ctx, NULL); + } + error = vn_close(vp, fg->fg_flag, ctx); + + (void)vnode_put(vp); + } + return(error); +} + +/* + * Returns: 0 Success + * VNOP_PATHCONF:??? */ int -vn_lock(vp, flags, p) - struct vnode *vp; - int flags; - struct proc *p; +vn_pathconf(vnode_t vp, int name, int32_t *retval, vfs_context_t ctx) +{ + int error = 0; + struct vfs_attr vfa; + + switch(name) { + case _PC_EXTENDED_SECURITY_NP: + *retval = vfs_extendedsecurity(vnode_mount(vp)) ? 1 : 0; + break; + case _PC_AUTH_OPAQUE_NP: + *retval = vfs_authopaque(vnode_mount(vp)); + break; + case _PC_2_SYMLINKS: + *retval = 1; /* XXX NOTSUP on MSDOS, etc. */ + break; + case _PC_ALLOC_SIZE_MIN: + *retval = 1; /* XXX lie: 1 byte */ + break; + case _PC_ASYNC_IO: /* unistd.h: _POSIX_ASYNCHRONUS_IO */ + *retval = 1; /* [AIO] option is supported */ + break; + case _PC_PRIO_IO: /* unistd.h: _POSIX_PRIORITIZED_IO */ + *retval = 0; /* [PIO] option is not supported */ + break; + case _PC_REC_INCR_XFER_SIZE: + *retval = 4096; /* XXX go from MIN to MAX 4K at a time */ + break; + case _PC_REC_MIN_XFER_SIZE: + *retval = 4096; /* XXX recommend 4K minimum reads/writes */ + break; + case _PC_REC_MAX_XFER_SIZE: + *retval = 65536; /* XXX recommend 64K maximum reads/writes */ + break; + case _PC_REC_XFER_ALIGN: + *retval = 4096; /* XXX recommend page aligned buffers */ + break; + case _PC_SYMLINK_MAX: + *retval = 255; /* Minimum acceptable POSIX value */ + break; + case _PC_SYNC_IO: /* unistd.h: _POSIX_SYNCHRONIZED_IO */ + *retval = 0; /* [SIO] option is not supported */ + break; + case _PC_XATTR_SIZE_BITS: + /* The number of bits used to store maximum extended + * attribute size in bytes. For example, if the maximum + * attribute size supported by a file system is 128K, the + * value returned will be 18. However a value 18 can mean + * that the maximum attribute size can be anywhere from + * (256KB - 1) to 128KB. As a special case, the resource + * fork can have much larger size, and some file system + * specific extended attributes can have smaller and preset + * size; for example, Finder Info is always 32 bytes. + */ + memset(&vfa, 0, sizeof(vfa)); + VFSATTR_INIT(&vfa); + VFSATTR_WANTED(&vfa, f_capabilities); + if (vfs_getattr(vnode_mount(vp), &vfa, ctx) == 0 && + (VFSATTR_IS_SUPPORTED(&vfa, f_capabilities)) && + (vfa.f_capabilities.capabilities[VOL_CAPABILITIES_INTERFACES] & VOL_CAP_INT_EXTENDED_ATTR) && + (vfa.f_capabilities.valid[VOL_CAPABILITIES_INTERFACES] & VOL_CAP_INT_EXTENDED_ATTR)) { + /* Supports native extended attributes */ + error = VNOP_PATHCONF(vp, name, retval, ctx); + } else { + /* Number of bits used to represent the maximum size of + * extended attribute stored in an Apple Double file. + */ + *retval = AD_XATTR_SIZE_BITS; + } + break; + default: + error = VNOP_PATHCONF(vp, name, retval, ctx); + break; + } + + return (error); +} + +static int +vn_kqfilt_add(struct fileproc *fp, struct knote *kn, vfs_context_t ctx) { int error; + struct vnode *vp; - do { - if ((flags & LK_INTERLOCK) == 0) - simple_lock(&vp->v_interlock); - if (vp->v_flag & VXLOCK) { - while (vp->v_flag & VXLOCK) { - vp->v_flag |= VXWANT; - simple_unlock(&vp->v_interlock); - (void)tsleep((caddr_t)vp, PINOD, "vn_lock", 0); - } - error = ENOENT; - } else { - error = VOP_LOCK(vp, flags | LK_INTERLOCK, p); - if (error == 0) - return (error); + vp = (struct vnode *)fp->f_fglob->fg_data; + + /* + * Don't attach a knote to a dead vnode. + */ + if ((error = vget_internal(vp, 0, VNODE_NODEAD)) == 0) { + switch (kn->kn_filter) { + case EVFILT_READ: + case EVFILT_WRITE: + if (vnode_isfifo(vp)) { + /* We'll only watch FIFOs that use our fifofs */ + if (!(vp->v_fifoinfo && vp->v_fifoinfo->fi_readsock)) { + error = ENOTSUP; + } + + } else if (!vnode_isreg(vp)) { + if (vnode_ischr(vp) && + (error = spec_kqfilter(vp, kn)) == 0) { + /* claimed by a special device */ + vnode_put(vp); + return 0; + } + + error = EINVAL; + } + break; + case EVFILT_VNODE: + break; + default: + error = EINVAL; + } + + if (error) { + vnode_put(vp); + return error; } - flags &= ~LK_INTERLOCK; - } while (flags & LK_RETRY); + +#if CONFIG_MACF + error = mac_vnode_check_kqfilter(ctx, fp->f_fglob->fg_cred, kn, vp); + if (error) { + vnode_put(vp); + return error; + } +#endif + + kn->kn_hook = (void*)vp; + kn->kn_hookid = vnode_vid(vp); + kn->kn_fop = &vnode_filtops; + + vnode_lock(vp); + KNOTE_ATTACH(&vp->v_knotes, kn); + vnode_unlock(vp); + + /* Ask the filesystem to provide remove notifications, but ignore failure */ + VNOP_MONITOR(vp, 0, VNODE_MONITOR_BEGIN, (void*) kn, ctx); + + vnode_put(vp); + } + return (error); } +static void +filt_vndetach(struct knote *kn) +{ + vfs_context_t ctx = vfs_context_current(); + struct vnode *vp; + vp = (struct vnode *)kn->kn_hook; + if (vnode_getwithvid(vp, kn->kn_hookid)) + return; + + vnode_lock(vp); + KNOTE_DETACH(&vp->v_knotes, kn); + vnode_unlock(vp); + + /* + * Tell a (generally networked) filesystem that we're no longer watching + * If the FS wants to track contexts, it should still be using the one from + * the VNODE_MONITOR_BEGIN. + */ + VNOP_MONITOR(vp, 0, VNODE_MONITOR_END, (void*)kn, ctx); + vnode_put(vp); +} + + /* - * File table vnode close routine. + * Used for EVFILT_READ + * + * Takes only VFIFO or VREG. vnode is locked. We handle the "poll" case + * differently than the regular case for VREG files. If not in poll(), + * then we need to know current fileproc offset for VREG. + */ +static intptr_t +vnode_readable_data_count(vnode_t vp, off_t current_offset, int ispoll) +{ + if (vnode_isfifo(vp)) { +#if FIFO + int cnt; + int err = fifo_charcount(vp, &cnt); + if (err == 0) { + return (intptr_t)cnt; + } else +#endif + { + return (intptr_t)0; + } + } else if (vnode_isreg(vp)) { + if (ispoll) { + return (intptr_t)1; + } + + off_t amount; + amount = vp->v_un.vu_ubcinfo->ui_size - current_offset; + if (amount > (off_t)INTPTR_MAX) { + return INTPTR_MAX; + } else if (amount < (off_t)INTPTR_MIN) { + return INTPTR_MIN; + } else { + return (intptr_t)amount; + } + } else { + panic("Should never have an EVFILT_READ except for reg or fifo."); + return 0; + } +} + +/* + * Used for EVFILT_WRITE. + * + * For regular vnodes, we can always write (1). For named pipes, + * see how much space there is in the buffer. Nothing else is covered. + */ +static intptr_t +vnode_writable_space_count(vnode_t vp) +{ + if (vnode_isfifo(vp)) { +#if FIFO + long spc; + int err = fifo_freespace(vp, &spc); + if (err == 0) { + return (intptr_t)spc; + } else +#endif + { + return (intptr_t)0; + } + } else if (vnode_isreg(vp)) { + return (intptr_t)1; + } else { + panic("Should never have an EVFILT_READ except for reg or fifo."); + return 0; + } +} + +/* + * Determine whether this knote should be active + * + * This is kind of subtle. + * --First, notice if the vnode has been revoked: in so, override hint + * --EVFILT_READ knotes are checked no matter what the hint is + * --Other knotes activate based on hint. + * --If hint is revoke, set special flags and activate */ static int -vn_closefile(fp, p) - struct file *fp; - struct proc *p; +filt_vnode(struct knote *kn, long hint) { + vnode_t vp = (struct vnode *)kn->kn_hook; + int activate = 0; + long orig_hint = hint; + + if (0 == hint) { + vnode_lock(vp); + + if (vnode_getiocount(vp, kn->kn_hookid, VNODE_NODEAD | VNODE_WITHID) != 0) { + /* Is recycled */ + hint = NOTE_REVOKE; + } + } else { + lck_mtx_assert(&vp->v_lock, LCK_MTX_ASSERT_OWNED); + } + + /* Special handling for vnodes that are in recycle or already gone */ + if (NOTE_REVOKE == hint) { + kn->kn_flags |= (EV_EOF | EV_ONESHOT); + activate = 1; + + if ((kn->kn_filter == EVFILT_VNODE) && (kn->kn_sfflags & NOTE_REVOKE)) { + kn->kn_fflags |= NOTE_REVOKE; + } + } else { + switch(kn->kn_filter) { + case EVFILT_READ: + kn->kn_data = vnode_readable_data_count(vp, kn->kn_fp->f_fglob->fg_offset, (kn->kn_flags & EV_POLL)); + + if (kn->kn_data != 0) { + activate = 1; + } + break; + case EVFILT_WRITE: + kn->kn_data = vnode_writable_space_count(vp); + + if (kn->kn_data != 0) { + activate = 1; + } + break; + case EVFILT_VNODE: + /* Check events this note matches against the hint */ + if (kn->kn_sfflags & hint) { + kn->kn_fflags |= hint; /* Set which event occurred */ + } + if (kn->kn_fflags != 0) { + activate = 1; + } + break; + default: + panic("Invalid knote filter on a vnode!\n"); + } + } + + if (orig_hint == 0) { + /* + * Definitely need to unlock, may need to put + */ + if (hint == 0) { + vnode_put_locked(vp); + } + vnode_unlock(vp); + } - return (vn_close(((struct vnode *)fp->f_data), fp->f_flag, - fp->f_cred, p)); + return (activate); }