freebsd-skq/usr.bin/make/dir.c

/*-
 * Copyright (c) 1988, 1989, 1990, 1993
 *	The Regents of the University of California.  All rights reserved.
 * Copyright (c) 1988, 1989 by Adam de Boor
 * Copyright (c) 1989 by Berkeley Softworks
 * All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * Adam de Boor.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 * @(#)dir.c	8.2 (Berkeley) 1/2/94
 */

#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");

/*-
 * dir.c --
 *	Directory searching using wildcards and/or normal names...
 *	Used both for source wildcarding in the Makefile and for finding
 *	implicit sources.
 *
 * The interface for this module is:
 *	Dir_Init	Initialize the module.
 *
 *	Dir_HasWildcards Returns TRUE if the name given it needs to
 *			be wildcard-expanded.
 *
 *	Path_Expand	Given a pattern and a path, return a Lst of names
 *			which match the pattern on the search path.
 *
 *	Path_FindFile	Searches for a file on a given search path.
 *			If it exists, the entire path is returned.
 *			Otherwise NULL is returned.
 *
 *	Dir_MTime	Return the modification time of a node. The file
 *			is searched for along the default search path.
 *			The path and mtime fields of the node are filled in.
 *
 *	Path_AddDir	Add a directory to a search path.
 *
 *	Dir_MakeFlags	Given a search path and a command flag, create
 *			a string with each of the directories in the path
 *			preceded by the command flag and all of them
 *			separated by a space.
 *
 *	Dir_Destroy	Destroy an element of a search path. Frees up all
 *			things that can be freed for the element as long
 *			as the element is no longer referenced by any other
 *			search path.
 *
 *	Dir_ClearPath	Resets a search path to the empty list.
 *
 * For debugging:
 *	Dir_PrintDirectories	Print stats about the directory cache.
 */

#include <sys/types.h>
#include <sys/stat.h>
#include <dirent.h>
#include <err.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>

#include "arch.h"
#include "dir.h"
#include "globals.h"
#include "GNode.h"
#include "hash.h"
#include "lst.h"
#include "make.h"
#include "str.h"
#include "targ.h"
#include "util.h"

/*
 *	A search path consists of a list of Dir structures. A Dir structure
 *	has in it the name of the directory and a hash table of all the files
 *	in the directory. This is used to cut down on the number of system
 *	calls necessary to find implicit dependents and their like. Since
 *	these searches are made before any actions are taken, we need not
 *	worry about the directory changing due to creation commands. If this
 *	hampers the style of some makefiles, they must be changed.
 *
 *	A list of all previously-read directories is kept in the
 *	openDirectories list. This list is checked first before a directory
 *	is opened.
 *
 *	The need for the caching of whole directories is brought about by
 *	the multi-level transformation code in suff.c, which tends to search
 *	for far more files than regular make does. In the initial
 *	implementation, the amount of time spent performing "stat" calls was
 *	truly astronomical. The problem with hashing at the start is,
 *	of course, that pmake doesn't then detect changes to these directories
 *	during the course of the make. Three possibilities suggest themselves:
 *
 *	    1) just use stat to test for a file's existence. As mentioned
 *	       above, this is very inefficient due to the number of checks
 *	       engendered by the multi-level transformation code.
 *	    2) use readdir() and company to search the directories, keeping
 *	       them open between checks. I have tried this and while it
 *	       didn't slow down the process too much, it could severely
 *	       affect the amount of parallelism available as each directory
 *	       open would take another file descriptor out of play for
 *	       handling I/O for another job. Given that it is only recently
 *	       that UNIX OS's have taken to allowing more than 20 or 32
 *	       file descriptors for a process, this doesn't seem acceptable
 *	       to me.
 *	    3) record the mtime of the directory in the Dir structure and
 *	       verify the directory hasn't changed since the contents were
 *	       hashed. This will catch the creation or deletion of files,
 *	       but not the updating of files. However, since it is the
 *	       creation and deletion that is the problem, this could be
 *	       a good thing to do. Unfortunately, if the directory (say ".")
 *	       were fairly large and changed fairly frequently, the constant
 *	       rehashing could seriously degrade performance. It might be
 *	       good in such cases to keep track of the number of rehashes
 *	       and if the number goes over a (small) limit, resort to using
 *	       stat in its place.
 *
 *	An additional thing to consider is that pmake is used primarily
 *	to create C programs and until recently pcc-based compilers refused
 *	to allow you to specify where the resulting object file should be
 *	placed. This forced all objects to be created in the current
 *	directory. This isn't meant as a full excuse, just an explanation of
 *	some of the reasons for the caching used here.
 *
 *	One more note: the location of a target's file is only performed
 *	on the downward traversal of the graph and then only for terminal
 *	nodes in the graph. This could be construed as wrong in some cases,
 *	but prevents inadvertent modification of files when the "installed"
 *	directory for a file is provided in the search path.
 *
 *	Another data structure maintained by this module is an mtime
 *	cache used when the searching of cached directories fails to find
 *	a file. In the past, Path_FindFile would simply perform an access()
 *	call in such a case to determine if the file could be found using
 *	just the name given. When this hit, however, all that was gained
 *	was the knowledge that the file existed. Given that an access() is
 *	essentially a stat() without the copyout() call, and that the same
 *	filesystem overhead would have to be incurred in Dir_MTime, it made
 *	sense to replace the access() with a stat() and record the mtime
 *	in a cache for when Dir_MTime was actually called.
 */

typedef struct Dir {
	char	*name;		/* Name of directory */
	int	refCount;	/* No. of paths with this directory */
	int	hits;		/* No. of times a file has been found here */
	Hash_Table files;	/* Hash table of files in directory */
	TAILQ_ENTRY(Dir) link;	/* allDirs link */
} Dir;

/*
 * A path is a list of pointers to directories. These directories are
 * reference counted so a directory can be on more than one path.
 */
struct PathElement {
	struct Dir	*dir;	/* pointer to the directory */
	TAILQ_ENTRY(PathElement) link;	/* path link */
};

/* main search path */
struct Path dirSearchPath = TAILQ_HEAD_INITIALIZER(dirSearchPath);

/* the list of all open directories */
static TAILQ_HEAD(, Dir) openDirectories =
    TAILQ_HEAD_INITIALIZER(openDirectories);

/*
 * Variables for gathering statistics on the efficiency of the hashing
 * mechanism.
 */
static int hits;	/* Found in directory cache */
static int misses;      /* Sad, but not evil misses */
static int nearmisses;	/* Found under search path */
static int bigmisses;	/* Sought by itself */

static Dir *dot;	    /* contents of current directory */

/* Results of doing a last-resort stat in Path_FindFile --
 * if we have to go to the system to find the file, we might as well
 * have its mtime on record.
 * XXX: If this is done way early, there's a chance other rules will
 * have already updated the file, in which case we'll update it again.
 * Generally, there won't be two rules to update a single file, so this
 * should be ok, but...
 */
static Hash_Table mtimes;

/*-
 *-----------------------------------------------------------------------
 * Dir_Init --
 *	initialize things for this module
 *
 * Results:
 *	none
 *
 * Side Effects:
 *	none
 *-----------------------------------------------------------------------
 */
void
Dir_Init(void)
{

	Hash_InitTable(&mtimes, 0);
}

/*-
 *-----------------------------------------------------------------------
 * Dir_InitDot --
 *	initialize the "." directory
 *
 * Results:
 *	none
 *
 * Side Effects:
 *	some directories may be opened.
 *-----------------------------------------------------------------------
 */
void
Dir_InitDot(void)
{

	dot = Path_AddDir(NULL, ".");
	if (dot == NULL)
		err(1, "cannot open current directory");

	/*
	 * We always need to have dot around, so we increment its
	 * reference count to make sure it's not destroyed.
	 */
	dot->refCount += 1;
}

/*-
 *-----------------------------------------------------------------------
 * Dir_HasWildcards  --
 *	See if the given name has any wildcard characters in it.
 *
 * Results:
 *	returns TRUE if the word should be expanded, FALSE otherwise
 *
 * Side Effects:
 *	none
 *-----------------------------------------------------------------------
 */
Boolean
Dir_HasWildcards(const char *name)
{
	const char *cp;
	int wild = 0, brace = 0, bracket = 0;

	for (cp = name; *cp; cp++) {
		switch (*cp) {
		case '{':
			brace++;
			wild = 1;
			break;
		case '}':
			brace--;
			break;
		case '[':
			bracket++;
			wild = 1;
			break;
		case ']':
			bracket--;
			break;
		case '?':
		case '*':
			wild = 1;
			break;
		default:
			break;
		}
	}
	return (wild && bracket == 0 && brace == 0);
}

/*-
 *-----------------------------------------------------------------------
 * DirMatchFiles --
 *	Given a pattern and a Dir structure, see if any files
 *	match the pattern and add their names to the 'expansions' list if
 *	any do. This is incomplete -- it doesn't take care of patterns like
 *	src / *src / *.c properly (just *.c on any of the directories), but it
 *	will do for now.
 *
 * Results:
 *	Always returns 0
 *
 * Side Effects:
 *	File names are added to the expansions lst. The directory will be
 *	fully hashed when this is done.
 *-----------------------------------------------------------------------
 */
static int
DirMatchFiles(const char *pattern, const Dir *p, Lst *expansions)
{
	Hash_Search search;	/* Index into the directory's table */
	Hash_Entry *entry;	/* Current entry in the table */
	Boolean isDot;		/* TRUE if the directory being searched is . */

	isDot = (*p->name == '.' && p->name[1] == '\0');

	for (entry = Hash_EnumFirst(&p->files, &search);
	    entry != NULL;
	    entry = Hash_EnumNext(&search)) {
		/*
		 * See if the file matches the given pattern. Note we follow
		 * the UNIX convention that dot files will only be found if
		 * the pattern begins with a dot (note also that as a side
		 * effect of the hashing scheme, .* won't match . or ..
		 * since they aren't hashed).
		 */
		if (Str_Match(entry->name, pattern) &&
		    ((entry->name[0] != '.') ||
		    (pattern[0] == '.'))) {
			Lst_AtEnd(expansions, (isDot ? estrdup(entry->name) :
			    str_concat(p->name, entry->name, STR_ADDSLASH)));
		}
	}
	return (0);
}

/*-
 *-----------------------------------------------------------------------
 * DirExpandCurly --
 *	Expand curly braces like the C shell. Does this recursively.
 *	Note the special case: if after the piece of the curly brace is
 *	done there are no wildcard characters in the result, the result is
 *	placed on the list WITHOUT CHECKING FOR ITS EXISTENCE.  The
 *	given arguments are the entire word to expand, the first curly
 *	brace in the word, the search path, and the list to store the
 *	expansions in.
 *
 * Results:
 *	None.
 *
 * Side Effects:
 *	The given list is filled with the expansions...
 *
 *-----------------------------------------------------------------------
 */
static void
DirExpandCurly(const char *word, const char *brace, struct Path *path,
    Lst *expansions)
{
	const char *end;	/* Character after the closing brace */
	const char *cp;		/* Current position in brace clause */
	const char *start;	/* Start of current piece of brace clause */
	int bracelevel;	/* Number of braces we've seen. If we see a right brace
			 * when this is 0, we've hit the end of the clause. */
	char *file;	/* Current expansion */
	int otherLen;	/* The length of the other pieces of the expansion
			 * (chars before and after the clause in 'word') */
	char *cp2;	/* Pointer for checking for wildcards in
			 * expansion before calling Dir_Expand */

	start = brace + 1;

	/*
	 * Find the end of the brace clause first, being wary of nested brace
	 * clauses.
	 */
	for (end = start, bracelevel = 0; *end != '\0'; end++) {
		if (*end == '{')
			bracelevel++;
		else if ((*end == '}') && (bracelevel-- == 0))
			break;
	}
	if (*end == '\0') {
		Error("Unterminated {} clause \"%s\"", start);
		return;
	} else
		end++;

	otherLen = brace - word + strlen(end);

	for (cp = start; cp < end; cp++) {
		/*
		 * Find the end of this piece of the clause.
		 */
		bracelevel = 0;
		while (*cp != ',') {
			if (*cp == '{')
				bracelevel++;
			else if ((*cp == '}') && (bracelevel-- <= 0))
				break;
			cp++;
		}
		/*
		 * Allocate room for the combination and install the
		 * three pieces.
		 */
		file = emalloc(otherLen + cp - start + 1);
		if (brace != word)
			strncpy(file, word, brace - word);
		if (cp != start)
			strncpy(&file[brace - word], start, cp - start);
		strcpy(&file[(brace - word) + (cp - start)], end);

		/*
		 * See if the result has any wildcards in it. If we find one,
		 * call Dir_Expand right away, telling it to place the result
		 * on our list of expansions.
		 */
		for (cp2 = file; *cp2 != '\0'; cp2++) {
			switch (*cp2) {
			case '*':
			case '?':
			case '{':
			case '[':
				Path_Expand(file, path, expansions);
				goto next;
			default:
				break;
			}
		}
		if (*cp2 == '\0') {
			/*
			 * Hit the end w/o finding any wildcards, so stick
			 * the expansion on the end of the list.
			 */
			Lst_AtEnd(expansions, file);
		} else {
		next:
			free(file);
		}
		start = cp + 1;
	}
}

/*-
 *-----------------------------------------------------------------------
 * DirExpandInt --
 *	Internal expand routine. Passes through the directories in the
 *	path one by one, calling DirMatchFiles for each. NOTE: This still
 *	doesn't handle patterns in directories...  Works given a word to
 *	expand, a path to look in, and a list to store expansions in.
 *
 * Results:
 *	None.
 *
 * Side Effects:
 *	Things are added to the expansions list.
 *
 *-----------------------------------------------------------------------
 */
static void
DirExpandInt(const char *word, const struct Path *path, Lst *expansions)
{
	struct PathElement *pe;

	TAILQ_FOREACH(pe, path, link)
		DirMatchFiles(word, pe->dir, expansions);
}

/*-
 *-----------------------------------------------------------------------
 * Dir_Expand  --
 *	Expand the given word into a list of words by globbing it looking
 *	in the directories on the given search path.
 *
 * Results:
 *	A list of words consisting of the files which exist along the search
 *	path matching the given pattern is placed in expansions.
 *
 * Side Effects:
 *	Directories may be opened. Who knows?
 *-----------------------------------------------------------------------
 */
void
Path_Expand(char *word, struct Path *path, Lst *expansions)
{
	LstNode *ln;
	char *cp;

	DEBUGF(DIR, ("expanding \"%s\"...", word));

	cp = strchr(word, '{');
	if (cp != NULL)
		DirExpandCurly(word, cp, path, expansions);
	else {
		cp = strchr(word, '/');
		if (cp != NULL) {
			/*
			 * The thing has a directory component -- find the
			 * first wildcard in the string.
			 */
			for (cp = word; *cp != '\0'; cp++) {
				if (*cp == '?' || *cp == '[' ||
				    *cp == '*' || *cp == '{') {
					break;
				}
			}
			if (*cp == '{') {
				/*
				 * This one will be fun.
				 */
				DirExpandCurly(word, cp, path, expansions);
				return;
			} else if (*cp != '\0') {
				/*
				 * Back up to the start of the component
				 */
				char *dirpath;

				while (cp > word && *cp != '/')
					cp--;
				if (cp != word) {
					char sc;

					/*
					 * If the glob isn't in the first
					 * component, try and find all the
					 * components up to the one with a
					 * wildcard.
					 */
					sc = cp[1];
					cp[1] = '\0';
					dirpath = Path_FindFile(word, path);
					cp[1] = sc;
					/*
					 * dirpath is null if can't find the
					 * leading component
					 * XXX: Path_FindFile won't find internal
					 * components. i.e. if the path contains
					 * ../Etc/Object and we're looking for
					 * Etc, * it won't be found. Ah well.
					 * Probably not important.
					 */
					if (dirpath != NULL) {
						char *dp =
						    &dirpath[strlen(dirpath)
						    - 1];
						struct Path tp =
						    TAILQ_HEAD_INITIALIZER(tp);

						if (*dp == '/')
							*dp = '\0';
						Path_AddDir(&tp, dirpath);
						DirExpandInt(cp + 1, &tp,
						    expansions);
						Path_Clear(&tp);
					}
				} else {
					/*
					 * Start the search from the local
					 * directory
					 */
					DirExpandInt(word, path, expansions);
				}
			} else {
				/*
				 * Return the file -- this should never happen.
				 */
				DirExpandInt(word, path, expansions);
			}
		} else {
			/*
			 * First the files in dot
			 */
			DirMatchFiles(word, dot, expansions);

			/*
			 * Then the files in every other directory on the path.
			 */
			DirExpandInt(word, path, expansions);
		}
	}
	if (DEBUG(DIR)) {
		LST_FOREACH(ln, expansions)
			DEBUGF(DIR, ("%s ", (const char *)Lst_Datum(ln)));
		DEBUGF(DIR, ("\n"));
	}
}

/**
 * Path_FindFile
 *	Find the file with the given name along the given search path.
 *
 * Results:
 *	The path to the file or NULL. This path is guaranteed to be in a
 *	different part of memory than name and so may be safely free'd.
 *
 * Side Effects:
 *	If the file is found in a directory which is not on the path
 *	already (either 'name' is absolute or it is a relative path
 *	[ dir1/.../dirn/file ] which exists below one of the directories
 *	already on the search path), its directory is added to the end
 *	of the path on the assumption that there will be more files in
 *	that directory later on. Sometimes this is true. Sometimes not.
 */
char *
Path_FindFile(char *name, struct Path *path)
{
	char *p1;		/* pointer into p->name */
	char *p2;		/* pointer into name */
	char *file;		/* the current filename to check */
	const struct PathElement *pe;	/* current path member */
	char *cp;		/* final component of the name */
	Boolean hasSlash;	/* true if 'name' contains a / */
	struct stat stb;	/* Buffer for stat, if necessary */
	Hash_Entry *entry;	/* Entry for mtimes table */

	/*
	 * Find the final component of the name and note whether it has a
	 * slash in it (the name, I mean)
	 */
	cp = strrchr(name, '/');
	if (cp != NULL) {
		hasSlash = TRUE;
		cp += 1;
	} else {
		hasSlash = FALSE;
		cp = name;
	}

	DEBUGF(DIR, ("Searching for %s...", name));
	/*
	 * No matter what, we always look for the file in the current directory
	 * before anywhere else and we *do not* add the ./ to it if it exists.
	 * This is so there are no conflicts between what the user specifies
	 * (fish.c) and what pmake finds (./fish.c).
	 */
	if ((!hasSlash || (cp - name == 2 && *name == '.')) &&
	    (Hash_FindEntry(&dot->files, cp) != NULL)) {
		DEBUGF(DIR, ("in '.'\n"));
		hits += 1;
		dot->hits += 1;
		return (estrdup(name));
	}

	/*
	 * We look through all the directories on the path seeking one which
	 * contains the final component of the given name and whose final
	 * component(s) match the name's initial component(s). If such a beast
	 * is found, we concatenate the directory name and the final component
	 * and return the resulting string. If we don't find any such thing,
	 * we go on to phase two...
	 */
	TAILQ_FOREACH(pe, path, link) {
		DEBUGF(DIR, ("%s...", pe->dir->name));
		if (Hash_FindEntry(&pe->dir->files, cp) != NULL) {
			DEBUGF(DIR, ("here..."));
			if (hasSlash) {
				/*
				 * If the name had a slash, its initial
				 * components and p's final components must
				 * match. This is false if a mismatch is
				 * encountered before all of the initial
				 * components have been checked (p2 > name at
				 * the end of the loop), or we matched only
				 * part of one of the components of p
				 * along with all the rest of them (*p1 != '/').
				 */
				p1 = pe->dir->name + strlen(pe->dir->name) - 1;
				p2 = cp - 2;
				while (p2 >= name && p1 >= pe->dir->name &&
				    *p1 == *p2) {
					p1 -= 1; p2 -= 1;
				}
				if (p2 >= name || (p1 >= pe->dir->name &&
				    *p1 != '/')) {
					DEBUGF(DIR, ("component mismatch -- "
					    "continuing..."));
					continue;
				}
			}
			file = str_concat(pe->dir->name, cp, STR_ADDSLASH);
			DEBUGF(DIR, ("returning %s\n", file));
			pe->dir->hits += 1;
			hits += 1;
			return (file);
		} else if (hasSlash) {
			/*
			 * If the file has a leading path component and that
			 * component exactly matches the entire name of the
			 * current search directory, we assume the file
			 * doesn't exist and return NULL.
			 */
			for (p1 = pe->dir->name, p2 = name; *p1 && *p1 == *p2;
			    p1++, p2++)
				continue;
			if (*p1 == '\0' && p2 == cp - 1) {
				if (*cp == '\0' || ISDOT(cp) || ISDOTDOT(cp)) {
					DEBUGF(DIR, ("returning %s\n", name));
					return (estrdup(name));
				} else {
					DEBUGF(DIR, ("must be here but isn't --"
					    " returning NULL\n"));
					return (NULL);
				}
			}
		}
	}

	/*
	 * We didn't find the file on any existing members of the directory.
	 * If the name doesn't contain a slash, that means it doesn't exist.
	 * If it *does* contain a slash, however, there is still hope: it
	 * could be in a subdirectory of one of the members of the search
	 * path. (eg. /usr/include and sys/types.h. The above search would
	 * fail to turn up types.h in /usr/include, but it *is* in
	 * /usr/include/sys/types.h) If we find such a beast, we assume there
	 * will be more (what else can we assume?) and add all but the last
	 * component of the resulting name onto the search path (at the
	 * end). This phase is only performed if the file is *not* absolute.
	 */
	if (!hasSlash) {
		DEBUGF(DIR, ("failed.\n"));
		misses += 1;
		return (NULL);
	}

	if (*name != '/') {
		Boolean	checkedDot = FALSE;

		DEBUGF(DIR, ("failed. Trying subdirectories..."));
		TAILQ_FOREACH(pe, path, link) {
			if (pe->dir != dot) {
				file = str_concat(pe->dir->name,
				    name, STR_ADDSLASH);
			} else {
				/*
				 * Checking in dot -- DON'T put a leading ./
				 * on the thing.
				 */
				file = estrdup(name);
				checkedDot = TRUE;
			}
			DEBUGF(DIR, ("checking %s...", file));

			if (stat(file, &stb) == 0) {
				DEBUGF(DIR, ("got it.\n"));

				/*
				 * We've found another directory to search. We
				 * know there's a slash in 'file' because we put
				 * one there. We nuke it after finding it and
				 * call Path_AddDir to add this new directory
				 * onto the existing search path. Once that's
				 * done, we restore the slash and triumphantly
				 * return the file name, knowing that should a
				 * file in this directory every be referenced
				 * again in such a manner, we will find it
				 * without having to do numerous numbers of
				 * access calls. Hurrah!
				 */
				cp = strrchr(file, '/');
				*cp = '\0';
				Path_AddDir(path, file);
				*cp = '/';

				/*
				 * Save the modification time so if
				 * it's needed, we don't have to fetch it again.
				 */
				DEBUGF(DIR, ("Caching %s for %s\n",
				    Targ_FmtTime(stb.st_mtime), file));
				entry = Hash_CreateEntry(&mtimes, file,
				    (Boolean *)NULL);
				Hash_SetValue(entry,
				    (void *)(long)stb.st_mtime);
				nearmisses += 1;
				return (file);
			} else {
				free(file);
			}
		}

		DEBUGF(DIR, ("failed. "));

		if (checkedDot) {
			/*
			 * Already checked by the given name, since . was in
			 * the path, so no point in proceeding...
			 */
			DEBUGF(DIR, ("Checked . already, returning NULL\n"));
			return (NULL);
		}
	}

	/*
	 * Didn't find it that way, either. Sigh. Phase 3. Add its directory
	 * onto the search path in any case, just in case, then look for the
	 * thing in the hash table. If we find it, grand. We return a new
	 * copy of the name. Otherwise we sadly return a NULL pointer. Sigh.
	 * Note that if the directory holding the file doesn't exist, this will
	 * do an extra search of the final directory on the path. Unless
	 * something weird happens, this search won't succeed and life will
	 * be groovy.
	 *
	 * Sigh. We cannot add the directory onto the search path because
	 * of this amusing case:
	 * $(INSTALLDIR)/$(FILE): $(FILE)
	 *
	 * $(FILE) exists in $(INSTALLDIR) but not in the current one.
	 * When searching for $(FILE), we will find it in $(INSTALLDIR)
	 * b/c we added it here. This is not good...
	 */
#ifdef notdef
	cp[-1] = '\0';
	Path_AddDir(path, name);
	cp[-1] = '/';

	bigmisses += 1;
	pe = TAILQ_LAST(path, Path);
	if (pe == NULL)
		return (NULL);

	if (Hash_FindEntry(&pe->dir->files, cp) != NULL) {
		return (estrdup(name));

	return (NULL);
#else /* !notdef */
	DEBUGF(DIR, ("Looking for \"%s\"...", name));

	bigmisses += 1;
	entry = Hash_FindEntry(&mtimes, name);
	if (entry != NULL) {
		DEBUGF(DIR, ("got it (in mtime cache)\n"));
		return (estrdup(name));
	} else if (stat (name, &stb) == 0) {
		entry = Hash_CreateEntry(&mtimes, name, (Boolean *)NULL);
		DEBUGF(DIR, ("Caching %s for %s\n",
		    Targ_FmtTime(stb.st_mtime), name));
		Hash_SetValue(entry, (void *)(long)stb.st_mtime);
		return (estrdup(name));
	} else {
		DEBUGF(DIR, ("failed. Returning NULL\n"));
		return (NULL);
	}
#endif /* notdef */
}

/*-
 *-----------------------------------------------------------------------
 * Dir_MTime  --
 *	Find the modification time of the file described by gn along the
 *	search path dirSearchPath.
 *
 * Results:
 *	The modification time or 0 if it doesn't exist
 *
 * Side Effects:
 *	The modification time is placed in the node's mtime slot.
 *	If the node didn't have a path entry before, and Dir_FindFile
 *	found one for it, the full name is placed in the path slot.
 *-----------------------------------------------------------------------
 */
int
Dir_MTime(GNode *gn)
{
	char *fullName;		/* the full pathname of name */
	struct stat stb;	/* buffer for finding the mod time */
	Hash_Entry *entry;

	if (gn->type & OP_ARCHV)
		return (Arch_MTime(gn));

	else if (gn->path == NULL)
		fullName = Path_FindFile(gn->name, &dirSearchPath);
	else
		fullName = gn->path;

	if (fullName == NULL)
		fullName = estrdup(gn->name);

	entry = Hash_FindEntry(&mtimes, fullName);
	if (entry != NULL) {
		/*
		 * Only do this once -- the second time folks are checking to
		 * see if the file was actually updated, so we need to
		 * actually go to the filesystem.
		 */
		DEBUGF(DIR, ("Using cached time %s for %s\n",
		    Targ_FmtTime((time_t)(long)Hash_GetValue(entry)),
		    fullName));
		stb.st_mtime = (time_t)(long)Hash_GetValue(entry);
		Hash_DeleteEntry(&mtimes, entry);
	} else if (stat(fullName, &stb) < 0) {
		if (gn->type & OP_MEMBER) {
			if (fullName != gn->path)
				free(fullName);
			return (Arch_MemMTime(gn));
		} else {
			stb.st_mtime = 0;
		}
	}
	if (fullName && gn->path == (char *)NULL)
		gn->path = fullName;

	gn->mtime = stb.st_mtime;
	return (gn->mtime);
}

/*-
 *-----------------------------------------------------------------------
 * Path_AddDir --
 *	Add the given name to the end of the given path.
 *
 * Results:
 *	none
 *
 * Side Effects:
 *	A structure is added to the list and the directory is
 *	read and hashed.
 *-----------------------------------------------------------------------
 */
struct Dir *
Path_AddDir(struct Path *path, const char *name)
{
	Dir *d;			/* pointer to new Path structure */
	DIR *dir;		/* for reading directory */
	struct PathElement *pe;
	struct dirent *dp;	/* entry in directory */

	/* check whether we know this directory */
	TAILQ_FOREACH(d, &openDirectories, link) {
		if (strcmp(d->name, name) == 0) {
			/* Found it. */
			if (path == NULL)
				return (d);

			/* Check whether its already on the path. */
			TAILQ_FOREACH(pe, path, link) {
				if (pe->dir == d)
					return (d);
			}
			/* Add it to the path */
			d->refCount += 1;
			pe = emalloc(sizeof(*pe));
			pe->dir = d;
			TAILQ_INSERT_TAIL(path, pe, link);
			return (d);
		}
	}

	DEBUGF(DIR, ("Caching %s...", name));

	if ((dir = opendir(name)) == NULL) {
		DEBUGF(DIR, (" cannot open\n"));
		return (NULL);
	}

	d = emalloc(sizeof(*d));
	d->name = estrdup(name);
	d->hits = 0;
	d->refCount = 1;
	Hash_InitTable(&d->files, -1);

	while ((dp = readdir(dir)) != NULL) {
#if defined(sun) && defined(d_ino) /* d_ino is a sunos4 #define	for d_fileno */
		/*
		 * The sun directory library doesn't check for
		 * a 0 inode (0-inode slots just take up space),
		 * so we have to do it ourselves.
		 */
		if (dp->d_fileno == 0)
			continue;
#endif /* sun && d_ino */

		/* Skip the '.' and '..' entries by checking
		 * for them specifically instead of assuming
		 * readdir() reuturns them in that order when
		 * first going through a directory.  This is
		 * needed for XFS over NFS filesystems since
		 * SGI does not guarantee that these are the
		 * first two entries returned from readdir().
		 */
		if (ISDOT(dp->d_name) || ISDOTDOT(dp->d_name))
			continue;

		Hash_CreateEntry(&d->files, dp->d_name, (Boolean *)NULL);
	}
	closedir(dir);

	if (path != NULL) {
		/* Add it to the path */
		d->refCount += 1;
		pe = emalloc(sizeof(*pe));
		pe->dir = d;
		TAILQ_INSERT_TAIL(path, pe, link);
	}

	/* Add to list of all directories */
	TAILQ_INSERT_TAIL(&openDirectories, d, link);

	DEBUGF(DIR, ("done\n"));

	return (d);
}

/**
 * Path_Duplicate
 *	Duplicate a path. Ups the reference count for the directories.
 */
void
Path_Duplicate(struct Path *dst, const struct Path *src)
{
	struct PathElement *ped, *pes;

	TAILQ_FOREACH(pes, src, link) {
		ped = emalloc(sizeof(*ped));
		ped->dir = pes->dir;
		ped->dir->refCount++;
		TAILQ_INSERT_TAIL(dst, ped, link);
	}
}

/**
 * Path_MakeFlags
 *	Make a string by taking all the directories in the given search
 *	path and preceding them by the given flag. Used by the suffix
 *	module to create variables for compilers based on suffix search
 *	paths.
 *
 * Results:
 *	The string mentioned above. Note that there is no space between
 *	the given flag and each directory. The empty string is returned if
 *	Things don't go well.
 */
char *
Path_MakeFlags(const char *flag, const struct Path *path)
{
	char *str;	/* the string which will be returned */
	char *tstr;	/* the current directory preceded by 'flag' */
	char *nstr;
	const struct PathElement *pe;

	str = estrdup("");

	TAILQ_FOREACH(pe, path, link) {
		tstr = str_concat(flag, pe->dir->name, 0);
		nstr = str_concat(str, tstr, STR_ADDSPACE);
		free(str);
		free(tstr);
		str = nstr;
	}

	return (str);
}

/**
 * Path_Clear
 *
 *	Destroy a path. This decrements the reference counts of all
 *	directories of this path and, if a reference count goes 0,
 *	destroys the directory object.
 */
void
Path_Clear(struct Path *path)
{
	struct PathElement *pe;

	while ((pe = TAILQ_FIRST(path)) != NULL) {
		pe->dir->refCount--;
		TAILQ_REMOVE(path, pe, link);
		if (pe->dir->refCount == 0) {
			TAILQ_REMOVE(&openDirectories, pe->dir, link);
			Hash_DeleteTable(&pe->dir->files);
			free(pe->dir->name);
			free(pe->dir);
		}
		free(pe);
	}
}

/**
 * Path_Concat
 *
 *	Concatenate two paths, adding the second to the end of the first.
 *	Make sure to avoid duplicates.
 *
 * Side Effects:
 *	Reference counts for added dirs are upped.
 */
void
Path_Concat(struct Path *path1, const struct Path *path2)
{
	struct PathElement *p1, *p2;

	TAILQ_FOREACH(p2, path2, link) {
		TAILQ_FOREACH(p1, path1, link) {
			if (p1->dir == p2->dir)
				break;
		}
		if (p1 == NULL) {
			p1 = emalloc(sizeof(*p1));
			p1->dir = p2->dir;
			p1->dir->refCount++;
			TAILQ_INSERT_TAIL(path1, p1, link);
		}
	}
}

/********** DEBUG INFO **********/
void
Dir_PrintDirectories(void)
{
	const Dir *d;

	printf("#*** Directory Cache:\n");
	printf("# Stats: %d hits %d misses %d near misses %d losers (%d%%)\n",
	    hits, misses, nearmisses, bigmisses,
	    (hits + bigmisses + nearmisses ?
	    hits * 100 / (hits + bigmisses + nearmisses) : 0));
	printf("# %-20s referenced\thits\n", "directory");
	TAILQ_FOREACH(d, &openDirectories, link)
		printf("# %-20s %10d\t%4d\n", d->name, d->refCount, d->hits);
}

void
Path_Print(const struct Path *path)
{
	const struct PathElement *p;

	TAILQ_FOREACH(p, path, link)
		printf("%s ", p->dir->name);
}