/* $OpenBSD: dir.c,v 1.67 2015/01/23 22:35:57 espie Exp $ */ /* $NetBSD: dir.c,v 1.14 1997/03/29 16:51:26 christos Exp $ */ /* * Copyright (c) 1999 Marc Espie. * * Extensive code changes for the OpenBSD project. * * 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. * * THIS SOFTWARE IS PROVIDED BY THE OPENBSD PROJECT 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 OPENBSD * PROJECT 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. */ /* * Copyright (c) 1988, 1989, 1990 The Regents of the University of California. * 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. 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. */ #include #include #include #include #include #include #include #include #include #include "config.h" #include "defines.h" #include "dir.h" #include "lst.h" #include "memory.h" #include "buf.h" #include "gnode.h" #include "arch.h" #include "targ.h" #include "error.h" #include "str.h" #include "timestamp.h" /* A search path consists of a Lst of PathEntry structures. A Path * 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 * knownDirectories cache. * * 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 PathEntry 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, Dir_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. */ /* several data structures exist to handle caching of directory stuff. * * There is a global hash of directory names (knownDirectories), and each * read directory is kept there as one PathEntry instance. Such a structure * only contains the file names. * * There is a global hash of timestamps (modification times), so care must * be taken of giving the right file names to that structure. * * XXX A set of similar structure should exist at the Target level to properly * take care of VPATH issues. */ /* each directory is cached into a PathEntry structure. */ struct PathEntry { int refCount; /* ref-counted, can participate to * several paths */ struct ohash files; /* hash of name of files in the directory */ char name[1]; /* directory name */ }; /* PathEntry kept on knownDirectories */ static struct ohash_info dir_info = { offsetof(struct PathEntry, name), NULL, hash_calloc, hash_free, element_alloc }; static struct ohash knownDirectories; /* cache all open directories */ /* file names kept in a path entry */ static struct ohash_info file_info = { 0, NULL, hash_calloc, hash_free, element_alloc }; /* Global structure used to cache mtimes. XXX We don't cache an mtime * before a caller actually looks up for the given time, because of the * possibility a caller might update the file and invalidate the cache * entry, and we don't look up in this cache except as a last resort. */ struct file_stamp { struct timespec mtime; /* time stamp... */ char name[1]; /* ...for that file. */ }; static struct ohash mtimes; static struct ohash_info stamp_info = { offsetof(struct file_stamp, name), NULL, hash_calloc, hash_free, element_alloc }; static LIST theDefaultPath; /* main search path */ Lst defaultPath= &theDefaultPath; struct PathEntry *dot; /* contents of current directory */ /* add_file(path, name): add a file name to a path hash structure. */ static void add_file(struct PathEntry *, const char *); /* n = find_file_hashi(p, name, end, hv): retrieve name in a path hash * structure. */ static char *find_file_hashi(struct PathEntry *, const char *, const char *, uint32_t); /* stamp = find_stampi(name, end): look for (name, end) in the global * cache. */ static struct file_stamp *find_stampi(const char *, const char *); /* record_stamp(name, timestamp): record timestamp for name in the global * cache. */ static void record_stamp(const char *, struct timespec); static bool read_directory(struct PathEntry *); /* p = DirReaddiri(name, end): read an actual directory, caching results * as we go. */ static struct PathEntry *create_PathEntry(const char *, const char *); /* Debugging: show a dir name in a path. */ static void DirPrintDir(void *); /*** *** timestamp handling ***/ static void record_stamp(const char *file, struct timespec t) { unsigned int slot; const char *end = NULL; struct file_stamp *n; slot = ohash_qlookupi(&mtimes, file, &end); n = ohash_find(&mtimes, slot); if (n) n->mtime = t; else { n = ohash_create_entry(&stamp_info, file, &end); n->mtime = t; ohash_insert(&mtimes, slot, n); } } static struct file_stamp * find_stampi(const char *file, const char *efile) { return ohash_find(&mtimes, ohash_qlookupi(&mtimes, file, &efile)); } /*** *** PathEntry handling ***/ static void add_file(struct PathEntry *p, const char *file) { unsigned int slot; const char *end = NULL; char *n; struct ohash *h = &p->files; slot = ohash_qlookupi(h, file, &end); n = ohash_find(h, slot); if (n == NULL) { n = ohash_create_entry(&file_info, file, &end); ohash_insert(h, slot, n); } } static char * find_file_hashi(struct PathEntry *p, const char *file, const char *efile, uint32_t hv) { struct ohash *h = &p->files; return ohash_find(h, ohash_lookup_interval(h, file, efile, hv)); } static bool read_directory(struct PathEntry *p) { DIR *d; struct dirent *dp; if (DEBUG(DIR)) { printf("Caching %s...", p->name); fflush(stdout); } if ((d = opendir(p->name)) == NULL) return false; ohash_init(&p->files, 4, &file_info); while ((dp = readdir(d)) != NULL) { if (dp->d_name[0] == '.' && (dp->d_name[1] == '\0' || (dp->d_name[1] == '.' && dp->d_name[2] == '\0'))) continue; add_file(p, dp->d_name); } (void)closedir(d); if (DEBUG(DIR)) printf("done\n"); return true; } /* Read a directory, either from the disk, or from the cache. */ static struct PathEntry * create_PathEntry(const char *name, const char *ename) { struct PathEntry *p; unsigned int slot; slot = ohash_qlookupi(&knownDirectories, name, &ename); p = ohash_find(&knownDirectories, slot); if (p == NULL) { p = ohash_create_entry(&dir_info, name, &ename); p->refCount = 0; if (!read_directory(p)) { free(p); return NULL; } ohash_insert(&knownDirectories, slot, p); } p->refCount++; return p; } char * PathEntry_name(struct PathEntry *p) { return p->name; } /* Side Effects: cache the current directory */ void Dir_Init(void) { char *dotname = "."; Static_Lst_Init(defaultPath); ohash_init(&knownDirectories, 4, &dir_info); ohash_init(&mtimes, 4, &stamp_info); dot = create_PathEntry(dotname, dotname+1); if (!dot) Fatal("Can't access current directory"); } /*- *----------------------------------------------------------------------- * Dir_MatchFilesi -- * Given a pattern and a PathEntry 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. *----------------------------------------------------------------------- */ void Dir_MatchFilesi(const char *word, const char *eword, struct PathEntry *p, Lst expansions) { unsigned int search; /* Index into the directory's table */ const char *entry; /* Current entry in the table */ for (entry = ohash_first(&p->files, &search); entry != NULL; entry = ohash_next(&p->files, &search)) { /* See if the file matches the given pattern. We follow the UNIX * convention that dot files will only be found if the pattern * begins with a dot (the hashing scheme doesn't hash . or .., * so they won't match `.*'. */ if (*word != '.' && *entry == '.') continue; if (Str_Matchi(entry, strchr(entry, '\0'), word, eword)) Lst_AtEnd(expansions, p == dot ? estrdup(entry) : Str_concat(p->name, entry, '/')); } } /*- * 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. */ char * Dir_FindFileComplexi(const char *name, const char *ename, Lst path, bool checkCurdirFirst) { struct PathEntry *p; /* current path member */ char *p1; /* pointer into p->name */ const char *p2; /* pointer into name */ LstNode ln; /* a list element */ char *file; /* the current filename to check */ char *temp; /* index into file */ const char *basename; bool hasSlash; struct stat stb;/* Buffer for stat, if necessary */ struct file_stamp *entry; /* Entry for mtimes table */ uint32_t hv; /* hash value for last component in file name */ char *q; /* Str_dupi(name, ename) */ /* Find the final component of the name and note whether name has a * slash in it */ basename = Str_rchri(name, ename, '/'); if (basename) { hasSlash = true; basename++; } else { hasSlash = false; basename = name; } hv = ohash_interval(basename, &ename); if (DEBUG(DIR)) printf("Searching for %s...", name); /* Unless checkCurDirFirst is false, we always look for * the file in the current directory before anywhere else * and we always return exactly what the caller specified. */ if (checkCurdirFirst && (!hasSlash || (basename - name == 2 && *name == '.')) && find_file_hashi(dot, basename, ename, hv) != NULL) { if (DEBUG(DIR)) printf("in '.'\n"); return Str_dupi(name, ename); } /* Then, we look through all the directories on path, seeking one * containing the final component of name and whose final * component(s) match name's initial component(s). * If found, we concatenate the directory name and the * final component and return the resulting string. */ for (ln = Lst_First(path); ln != NULL; ln = Lst_Adv(ln)) { p = (struct PathEntry *)Lst_Datum(ln); if (DEBUG(DIR)) printf("%s...", p->name); if (find_file_hashi(p, basename, ename, hv) != NULL) { if (DEBUG(DIR)) printf("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 = p->name + strlen(p->name) - 1; p2 = basename - 2; while (p2 >= name && p1 >= p->name && *p1 == *p2) { p1--; p2--; } if (p2 >= name || (p1 >= p->name && *p1 != '/')) { if (DEBUG(DIR)) printf("component mismatch -- continuing..."); continue; } } file = Str_concati(p->name, strchr(p->name, '\0'), basename, ename, '/'); if (DEBUG(DIR)) printf("returning %s\n", file); 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 = p->name, p2 = name; *p1 && *p1 == *p2; p1++, p2++) continue; if (*p1 == '\0' && p2 == basename - 1) { if (DEBUG(DIR)) printf("has to be here but isn't -- returning NULL\n"); return NULL; } } } /* We didn't find the file on any existing member of the path. * If the name doesn't contain a slash, end of story. * If it does contain a slash, however, it could be in a subdirectory * of one of the members of the search path. (eg., for path=/usr/include * and name=sys/types.h, the above search fails to turn up types.h * in /usr/include, even though /usr/include/sys/types.h exists). * * We only perform this look-up for non-absolute file names. * * Whenever we score a hit, we assume there will be more matches from * that directory, and append all but the last component of the * resulting name onto the search path. */ if (!hasSlash) { if (DEBUG(DIR)) printf("failed.\n"); return NULL; } if (*name != '/') { bool checkedDot = false; if (DEBUG(DIR)) printf("failed. Trying subdirectories..."); for (ln = Lst_First(path); ln != NULL; ln = Lst_Adv(ln)) { p = (struct PathEntry *)Lst_Datum(ln); if (p != dot) file = Str_concati(p->name, strchr(p->name, '\0'), name, ename, '/'); else { /* Checking in dot -- DON'T put a leading * ./ on the thing. */ file = Str_dupi(name, ename); checkedDot = true; } if (DEBUG(DIR)) printf("checking %s...", file); if (stat(file, &stb) == 0) { struct timespec mtime; ts_set_from_stat(stb, mtime); if (DEBUG(DIR)) printf("got it.\n"); /* We've found another directory to search. * We know there is a slash in 'file'. We * call Dir_AddDiri to add the new directory * onto the existing search path. Once that's * done, we return the file name, knowing that * should a file in this directory ever be * referenced again in such a manner, we will * find it without having to do numerous * access calls. */ temp = strrchr(file, '/'); Dir_AddDiri(path, file, temp); /* Save the modification time so if it's * needed, we don't have to fetch it again. */ if (DEBUG(DIR)) printf("Caching %s for %s\n", time_to_string(&mtime), file); record_stamp(file, mtime); return file; } else free(file); } if (DEBUG(DIR)) printf("failed. "); if (checkedDot) { /* Already checked by the given name, since . was in * the path, so no point in proceeding... */ if (DEBUG(DIR)) printf("Checked . already, returning NULL\n"); return NULL; } } /* Didn't find it that way, either. Last resort: look for the file * in the global mtime cache, then on the disk. * If this doesn't succeed, we finally return a NULL pointer. * * We cannot add this 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... */ q = Str_dupi(name, ename); if (DEBUG(DIR)) printf("Looking for \"%s\"...", q); entry = find_stampi(name, ename); if (entry != NULL) { if (DEBUG(DIR)) printf("got it (in mtime cache)\n"); return q; } else if (stat(q, &stb) == 0) { struct timespec mtime; ts_set_from_stat(stb, mtime); if (DEBUG(DIR)) printf("Caching %s for %s\n", time_to_string(&mtime), q); record_stamp(q, mtime); return q; } else { if (DEBUG(DIR)) printf("failed. Returning NULL\n"); free(q); return NULL; } } void Dir_AddDiri(Lst path, const char *name, const char *ename) { struct PathEntry *p; p = create_PathEntry(name, ename); if (p == NULL) return; if (p->refCount == 1) Lst_AtEnd(path, p); else if (!Lst_AddNew(path, p)) return; } void * Dir_CopyDir(void *p) { struct PathEntry *q = p; q->refCount++; return p; } void Dir_Destroy(void *pp) { struct PathEntry *p = pp; if (--p->refCount == 0) { ohash_remove(&knownDirectories, ohash_qlookup(&knownDirectories, p->name)); free_hash(&p->files); free(p); } } /*- *----------------------------------------------------------------------- * Dir_Concat -- * Concatenate two paths, adding the second to the end of the first. * Makes sure to avoid duplicates. * * Side Effects: * Reference counts for added dirs are upped. *----------------------------------------------------------------------- */ void Dir_Concat(Lst path1, Lst path2) { LstNode ln; struct PathEntry *p; for (ln = Lst_First(path2); ln != NULL; ln = Lst_Adv(ln)) { p = (struct PathEntry *)Lst_Datum(ln); if (Lst_AddNew(path1, p)) p->refCount++; } } static void DirPrintDir(void *p) { const struct PathEntry *q = p; printf("%s ", q->name); } void Dir_PrintPath(Lst path) { Lst_Every(path, DirPrintDir); } struct timespec Dir_MTime(GNode *gn) { char *fullName; struct stat stb; struct file_stamp *entry; unsigned int slot; struct timespec mtime; if (gn->type & OP_PHONY) return gn->mtime; if (gn->type & OP_ARCHV) return Arch_MTime(gn); if (gn->path == NULL) { fullName = Dir_FindFile(gn->name, defaultPath); if (fullName == NULL) fullName = estrdup(gn->name); } else fullName = gn->path; slot = ohash_qlookup(&mtimes, fullName); entry = ohash_find(&mtimes, slot); 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 file system. */ if (DEBUG(DIR)) printf("Using cached time %s for %s\n", time_to_string(&entry->mtime), fullName); mtime = entry->mtime; free(entry); ohash_remove(&mtimes, slot); } else if (stat(fullName, &stb) == 0) ts_set_from_stat(stb, mtime); else { if (gn->type & OP_MEMBER) { if (fullName != gn->path) free(fullName); return Arch_MemMTime(gn); } else ts_set_out_of_date(mtime); } if (fullName && gn->path == NULL) gn->path = fullName; gn->mtime = mtime; return gn->mtime; }