Path handling related refactorings

Primarily related to the path handling datatypes, to avoid unnecessary
casts, where possible. Also some rework to avoid code dup. Probably
more places are to go, even not path related, primarily to have less
casts and unsigned integers where possible. That way, we've not only
less warnings and casts, but are also safer with regard to the
integer overflows. OFC it's not a panacea, but still significantly
reduces the vulnerability potential.

13 files changed, 54 insertions, 97 deletions

diff --git a/TSRM/readdir.h b/TSRM/readdir.h
index 0665810c85..11c0f31ea1 100644
--- a/TSRM/readdir.h
+++ b/TSRM/readdir.h
@@ -1,45 +1,3 @@
-#ifndef READDIR_H
-#define READDIR_H
-
-
-/*
- * Structures and types used to implement opendir/readdir/closedir
- * on Windows 95/NT.
- */
-
-#include <windows.h>
-
-#include <io.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include <sys/types.h>
-#include <direct.h>
-
-/* struct dirent - same as Unix */
-
-struct dirent {
-	long d_ino;					/* inode (always 1 in WIN32) */
-	off_t d_off;				/* offset to this dirent */
-	unsigned short d_reclen;	/* length of d_name */
-	char d_name[_MAX_FNAME + 1];	/* filename (null terminated) */
-};
-
-
-/* typedef DIR - not the same as Unix */
-typedef struct {
-	HANDLE handle;				/* _findfirst/_findnext handle */
-	short offset;				/* offset into directory */
-	short finished;				/* 1 if there are not more files */
-	WIN32_FIND_DATA fileinfo;	/* from _findfirst/_findnext */
-	char *dir;					/* the dir we are reading */
-	struct dirent dent;			/* the dirent to return */
-} DIR;
-
-/* Function prototypes */
-DIR *opendir(const char *);
-struct dirent *readdir(DIR *);
-int readdir_r(DIR *, struct dirent *, struct dirent **);
-int closedir(DIR *);
-int rewinddir(DIR *);
-
-#endif /* READDIR_H */
+/* Keep this header for compatibility with external code, it's currently not
+	used anywhere in the core and there are no implementations in TSRM. */
+#include "win32/readdir.h"
diff --git a/TSRM/tsrm_win32.c b/TSRM/tsrm_win32.c
index 7fa38198de..ac9d667788 100644
--- a/TSRM/tsrm_win32.c
+++ b/TSRM/tsrm_win32.c
@@ -296,14 +296,14 @@ TSRM_API int tsrm_win32_access(const char *pathname, int mode)
 
 		if (CWDG(realpath_cache_size_limit)) {
 			t = time(0);
-			bucket = realpath_cache_lookup(pathname, (int)strlen(pathname), t);
+			bucket = realpath_cache_lookup(pathname, strlen(pathname), t);
 			if(bucket == NULL && real_path == NULL) {
 				/* We used the pathname directly. Call tsrm_realpath */
 				/* so that entry is created in realpath cache */
 				real_path = (char *)malloc(MAXPATHLEN);
 				if(tsrm_realpath(pathname, real_path) != NULL) {
 					pathname = real_path;
-					bucket = realpath_cache_lookup(pathname, (int)strlen(pathname), t);
+					bucket = realpath_cache_lookup(pathname, strlen(pathname), t);
 					PHP_WIN32_IOUTIL_REINIT_W(pathname);
 				}
 			}
diff --git a/UPGRADING.INTERNALS b/UPGRADING.INTERNALS
index 95e9318b25..a9f1547245 100644
--- a/UPGRADING.INTERNALS
+++ b/UPGRADING.INTERNALS
@@ -13,7 +13,12 @@ PHP 7.1 INTERNALS UPGRADE NOTES
 1. Internal API changes
 ========================
 
-  a.
+  a. Path related functions
+    - CWD_API void realpath_cache_del(const char *path, size_t path_len);
+    - CWD_API realpath_cache_bucket* realpath_cache_lookup(const char *path, size_t path_len, time_t t);
+    - PHPAPI void php_clear_stat_cache(zend_bool clear_realpath_cache, const char *filename, size_t filename_len);
+    - PHPAPI void php_stat(const char *filename, size_t filename_length, int type, zval *return_value);
+
 
 ========================
 2. Build system changes
diff --git a/Zend/zend_virtual_cwd.c b/Zend/zend_virtual_cwd.c
index 420bc54a7b..02f3834273 100644
--- a/Zend/zend_virtual_cwd.c
+++ b/Zend/zend_virtual_cwd.c
@@ -537,7 +537,7 @@ CWD_API char *virtual_getcwd(char *buf, size_t size) /* {{{ */
 /* }}} */
 
 #ifdef ZEND_WIN32
-static inline zend_ulong realpath_cache_key(const char *path, int path_len) /* {{{ */
+static inline zend_ulong realpath_cache_key(const char *path, size_t path_len) /* {{{ */
 {
 	register zend_ulong h;
 	char *bucket_key_start = tsrm_win32_get_path_sid_key(path);
@@ -558,7 +558,7 @@ static inline zend_ulong realpath_cache_key(const char *path, int path_len) /* {
 }
 /* }}} */
 #else
-static inline zend_ulong realpath_cache_key(const char *path, int path_len) /* {{{ */
+static inline zend_ulong realpath_cache_key(const char *path, size_t path_len) /* {{{ */
 {
 	register zend_ulong h;
 	const char *e = path + path_len;
@@ -590,7 +590,7 @@ CWD_API void realpath_cache_clean(void) /* {{{ */
 }
 /* }}} */
 
-CWD_API void realpath_cache_del(const char *path, int path_len) /* {{{ */
+CWD_API void realpath_cache_del(const char *path, size_t path_len) /* {{{ */
 {
 	zend_ulong key = realpath_cache_key(path, path_len);
 	zend_ulong n = key % (sizeof(CWDG(realpath_cache)) / sizeof(CWDG(realpath_cache)[0]));
@@ -618,7 +618,7 @@ CWD_API void realpath_cache_del(const char *path, int path_len) /* {{{ */
 }
 /* }}} */
 
-static inline void realpath_cache_add(const char *path, int path_len, const char *realpath, int realpath_len, int is_dir, time_t t) /* {{{ */
+static inline void realpath_cache_add(const char *path, int path_len, const char *realpath, size_t realpath_len, int is_dir, time_t t) /* {{{ */
 {
 	zend_long size = sizeof(realpath_cache_bucket) + path_len + 1;
 	int same = 1;
@@ -664,7 +664,7 @@ static inline void realpath_cache_add(const char *path, int path_len, const char
 }
 /* }}} */
 
-static inline realpath_cache_bucket* realpath_cache_find(const char *path, int path_len, time_t t) /* {{{ */
+static inline realpath_cache_bucket* realpath_cache_find(const char *path, size_t path_len, time_t t) /* {{{ */
 {
 	zend_ulong key = realpath_cache_key(path, path_len);
 	zend_ulong n = key % (sizeof(CWDG(realpath_cache)) / sizeof(CWDG(realpath_cache)[0]));
@@ -693,7 +693,7 @@ static inline realpath_cache_bucket* realpath_cache_find(const char *path, int p
 }
 /* }}} */
 
-CWD_API realpath_cache_bucket* realpath_cache_lookup(const char *path, int path_len, time_t t) /* {{{ */
+CWD_API realpath_cache_bucket* realpath_cache_lookup(const char *path, size_t path_len, time_t t) /* {{{ */
 {
 	return realpath_cache_find(path, path_len, t);
 }
diff --git a/Zend/zend_virtual_cwd.h b/Zend/zend_virtual_cwd.h
index 970e65e0b7..74e9f569fd 100644
--- a/Zend/zend_virtual_cwd.h
+++ b/Zend/zend_virtual_cwd.h
@@ -232,8 +232,8 @@ extern virtual_cwd_globals cwd_globals;
 #endif
 
 CWD_API void realpath_cache_clean(void);
-CWD_API void realpath_cache_del(const char *path, int path_len);
-CWD_API realpath_cache_bucket* realpath_cache_lookup(const char *path, int path_len, time_t t);
+CWD_API void realpath_cache_del(const char *path, size_t path_len);
+CWD_API realpath_cache_bucket* realpath_cache_lookup(const char *path, size_t path_len, time_t t);
 CWD_API zend_long realpath_cache_size(void);
 CWD_API zend_long realpath_cache_max_buckets(void);
 CWD_API realpath_cache_bucket** realpath_cache_get_buckets(void);
diff --git a/ext/phar/func_interceptors.c b/ext/phar/func_interceptors.c
index 05bcbf9c36..3e78fa1f86 100644
--- a/ext/phar/func_interceptors.c
+++ b/ext/phar/func_interceptors.c
@@ -585,7 +585,7 @@ static void phar_fancy_stat(zend_stat_t *stat_sb, int type, zval *return_value)
 }
 /* }}} */
 
-static void phar_file_stat(const char *filename, php_stat_len filename_length, int type, void (*orig_stat_func)(INTERNAL_FUNCTION_PARAMETERS), INTERNAL_FUNCTION_PARAMETERS) /* {{{ */
+static void phar_file_stat(const char *filename, size_t filename_length, int type, void (*orig_stat_func)(INTERNAL_FUNCTION_PARAMETERS), INTERNAL_FUNCTION_PARAMETERS) /* {{{ */
 {
 	if (!filename_length) {
 		RETURN_FALSE;
@@ -770,7 +770,7 @@ void fname(INTERNAL_FUNCTION_PARAMETERS) { \
 			return; \
 		} \
 		\
-		phar_file_stat(filename, (php_stat_len) filename_len, funcnum, PHAR_G(orig), INTERNAL_FUNCTION_PARAM_PASSTHRU); \
+		phar_file_stat(filename, filename_len, funcnum, PHAR_G(orig), INTERNAL_FUNCTION_PARAM_PASSTHRU); \
 	} \
 }
 /* }}} */
diff --git a/ext/phar/phar_object.c b/ext/phar/phar_object.c
index 457d6e593d..8af3869d9a 100644
--- a/ext/phar/phar_object.c
+++ b/ext/phar/phar_object.c
@@ -1423,7 +1423,7 @@ static int phar_build(zend_object_iterator *iter, void *puser) /* {{{ */
 	uint32_t str_key_len, base_len = p_obj->l;
 	phar_entry_data *data;
 	php_stream *fp;
-	php_stat_len fname_len;
+	size_t fname_len;
 	size_t contents_len;
 	char *fname, *error = NULL, *base = p_obj->b, *save = NULL, *temp = NULL;
 	zend_string *opened;
@@ -1501,7 +1501,7 @@ static int phar_build(zend_object_iterator *iter, void *puser) /* {{{ */
 				switch (intern->type) {
 					case SPL_FS_DIR:
 						test = spl_filesystem_object_get_path(intern, NULL);
-						fname_len = (php_stat_len)spprintf(&fname, 0, "%s%c%s", test, DEFAULT_SLASH, intern->u.dir.entry.d_name);
+						fname_len = spprintf(&fname, 0, "%s%c%s", test, DEFAULT_SLASH, intern->u.dir.entry.d_name);
 						php_stat(fname, fname_len, FS_IS_DIR, &dummy);
 
 						if (Z_TYPE(dummy) == IS_TRUE) {
@@ -1515,7 +1515,7 @@ static int phar_build(zend_object_iterator *iter, void *puser) /* {{{ */
 
 						if (test) {
 							fname = test;
-							fname_len = (php_stat_len)strlen(fname);
+							fname_len = strlen(fname);
 						} else {
 							zend_throw_exception_ex(spl_ce_UnexpectedValueException, 0, "Could not resolve file path");
 							return ZEND_HASH_APPLY_STOP;
@@ -1531,7 +1531,7 @@ static int phar_build(zend_object_iterator *iter, void *puser) /* {{{ */
 							return ZEND_HASH_APPLY_STOP;
 						}
 
-						fname_len = (php_stat_len)strlen(fname);
+						fname_len = strlen(fname);
 						save = fname;
 						goto phar_spl_fileinfo;
 				}
@@ -1543,7 +1543,7 @@ static int phar_build(zend_object_iterator *iter, void *puser) /* {{{ */
 	}
 
 	fname = Z_STRVAL_P(value);
-	fname_len = (php_stat_len)Z_STRLEN_P(value);
+	fname_len = Z_STRLEN_P(value);
 
 phar_spl_fileinfo:
 	if (base_len) {
diff --git a/ext/standard/dir.c b/ext/standard/dir.c
index de7e01ffea..9e7adfccfb 100644
--- a/ext/standard/dir.c
+++ b/ext/standard/dir.c
@@ -420,7 +420,7 @@ PHP_NAMED_FUNCTION(php_if_readdir)
    Find pathnames matching a pattern */
 PHP_FUNCTION(glob)
 {
-	int cwd_skip = 0;
+	size_t cwd_skip = 0;
 #ifdef ZTS
 	char cwd[MAXPATHLEN];
 	char work_pattern[MAXPATHLEN];
@@ -459,7 +459,7 @@ PHP_FUNCTION(glob)
 			cwd[2] = '\0';
 		}
 #endif
-		cwd_skip = (int)strlen(cwd)+1;
+		cwd_skip = strlen(cwd)+1;
 
 		snprintf(work_pattern, MAXPATHLEN, "%s%c%s", cwd, DEFAULT_SLASH, pattern);
 		pattern = work_pattern;
diff --git a/ext/standard/filestat.c b/ext/standard/filestat.c
index 370805a27e..47e7d63738 100644
--- a/ext/standard/filestat.c
+++ b/ext/standard/filestat.c
@@ -705,7 +705,7 @@ PHP_FUNCTION(touch)
 
 /* {{{ php_clear_stat_cache()
 */
-PHPAPI void php_clear_stat_cache(zend_bool clear_realpath_cache, const char *filename, int filename_len)
+PHPAPI void php_clear_stat_cache(zend_bool clear_realpath_cache, const char *filename, size_t filename_len)
 {
 	/* always clear CurrentStatFile and CurrentLStatFile even if filename is not NULL
 	 * as it may contain outdated data (e.g. "nlink" for a directory when deleting a file
@@ -740,7 +740,7 @@ PHP_FUNCTION(clearstatcache)
 		return;
 	}
 
-	php_clear_stat_cache(clear_realpath_cache, filename, (int)filename_len);
+	php_clear_stat_cache(clear_realpath_cache, filename, filename_len);
 }
 /* }}} */
 
@@ -751,7 +751,7 @@ PHP_FUNCTION(clearstatcache)
 
 /* {{{ php_stat
  */
-PHPAPI void php_stat(const char *filename, php_stat_len filename_length, int type, zval *return_value)
+PHPAPI void php_stat(const char *filename, size_t filename_length, int type, zval *return_value)
 {
 	zval stat_dev, stat_ino, stat_mode, stat_nlink, stat_uid, stat_gid, stat_rdev,
 		 stat_size, stat_atime, stat_mtime, stat_ctime, stat_blksize, stat_blocks;
@@ -996,7 +996,7 @@ void name(INTERNAL_FUNCTION_PARAMETERS) { \
 		Z_PARAM_PATH(filename, filename_len) \
 	ZEND_PARSE_PARAMETERS_END(); \
 	\
-	php_stat(filename, (php_stat_len) filename_len, funcnum, return_value); \
+	php_stat(filename, filename_len, funcnum, return_value); \
 }
 /* }}} */
 
diff --git a/ext/standard/php_filestat.h b/ext/standard/php_filestat.h
index 2e87d96f05..5e141ca3e4 100644
--- a/ext/standard/php_filestat.h
+++ b/ext/standard/php_filestat.h
@@ -78,14 +78,11 @@ PHP_FUNCTION(clearstatcache);
 #define getuid() 1
 #endif
 
-#ifdef PHP_WIN32
-typedef unsigned int php_stat_len;
-#else
-typedef int php_stat_len;
-#endif
+/* Compatibility. */
+typedef size_t php_stat_len;
 
-PHPAPI void php_clear_stat_cache(zend_bool clear_realpath_cache, const char *filename, int filename_len);
-PHPAPI void php_stat(const char *filename, php_stat_len filename_length, int type, zval *return_value);
+PHPAPI void php_clear_stat_cache(zend_bool clear_realpath_cache, const char *filename, size_t filename_len);
+PHPAPI void php_stat(const char *filename, size_t filename_length, int type, zval *return_value);
 
 /* Switches for various filestat functions: */
 #define FS_PERMS    0
diff --git a/sapi/cli/php_cli.c b/sapi/cli/php_cli.c
index 65a0a63505..b5fa6ea989 100644
--- a/sapi/cli/php_cli.c
+++ b/sapi/cli/php_cli.c
@@ -1126,7 +1126,7 @@ static int do_cli(int argc, char **argv) /* {{{ */
 				}
 			case PHP_MODE_REFLECTION_EXT_INFO:
 				{
-					int len = (int)strlen(reflection_what);
+					size_t len = strlen(reflection_what);
 					char *lcname = zend_str_tolower_dup(reflection_what, len);
 					zend_module_entry *module;
 
@@ -1202,7 +1202,7 @@ int main(int argc, char *argv[])
 	int php_optind = 1, use_extended_info = 0;
 	char *ini_path_override = NULL;
 	char *ini_entries = NULL;
-	int ini_entries_len = 0;
+	size_t ini_entries_len = 0;
 	int ini_ignore = 0;
 	sapi_module_struct *sapi_module = &cli_sapi_module;
 
@@ -1276,7 +1276,7 @@ int main(int argc, char *argv[])
 				break;
 			case 'd': {
 				/* define ini entries on command line */
-				int len = (int)strlen(php_optarg);
+				size_t len = strlen(php_optarg);
 				char *val;
 
 				if ((val = strchr(php_optarg, '='))) {
@@ -1284,11 +1284,11 @@ int main(int argc, char *argv[])
 					if (!isalnum(*val) && *val != '"' && *val != '\'' && *val != '\0') {
 						ini_entries = realloc(ini_entries, ini_entries_len + len + sizeof("\"\"\n\0"));
 						memcpy(ini_entries + ini_entries_len, php_optarg, (val - php_optarg));
-						ini_entries_len += (int)(val - php_optarg);
+						ini_entries_len += (val - php_optarg);
 						memcpy(ini_entries + ini_entries_len, "\"", 1);
 						ini_entries_len++;
 						memcpy(ini_entries + ini_entries_len, val, len - (val - php_optarg));
-						ini_entries_len += len - (int)(val - php_optarg);
+						ini_entries_len += len - (val - php_optarg);
 						memcpy(ini_entries + ini_entries_len, "\"\n\0", sizeof("\"\n\0"));
 						ini_entries_len += sizeof("\n\0\"") - 2;
 					} else {
diff --git a/win32/readdir.c b/win32/readdir.c
index 4a6d65932f..43d5deecfd 100644
--- a/win32/readdir.c
+++ b/win32/readdir.c
@@ -24,16 +24,6 @@
 extern "C" {
 #endif
 
-/* typedef DIR - not the same as Unix */
-struct DIR_W32 {
-	HANDLE handle;				/* _findfirst/_findnext handle */
-	int offset;					/* offset into directory */
-	short finished;				/* 1 if there are not more files */
-	WIN32_FIND_DATAW fileinfo;  /* from _findfirst/_findnext */
-	wchar_t *dirw;		/* the dir we are reading */
-	struct dirent dent;			/* the dirent to return */
-};
-
 DIR *opendir(const char *dir)
 {
 	DIR *dp;
diff --git a/win32/readdir.h b/win32/readdir.h
index 4158ffc84a..73058f8804 100644
--- a/win32/readdir.h
+++ b/win32/readdir.h
@@ -8,25 +8,32 @@ extern "C" {
 
 /*
  * Structures and types used to implement opendir/readdir/closedir
- * on Windows 95/NT.
+ * on Windows.
  */
 
 #include <config.w32.h>
 
-#include <stdlib.h>
-#include <sys/types.h>
+#include "ioutil.h"
 
 #define php_readdir_r readdir_r
 
 /* struct dirent - same as Unix */
 struct dirent {
 	long d_ino;					/* inode (always 1 in WIN32) */
-	off_t d_off;				/* offset to this dirent */
-	unsigned short d_reclen;	/* length of d_name */
-	char d_name[_MAX_FNAME + 1];	/* filename (null terminated) */
+	off_t d_off;					/* offset to this dirent */
+	unsigned short d_reclen;			/* length of d_name */
+	char d_name[PHP_WIN32_IOUTIL_MAXPATHLEN + 1];	/* filename (null terminated) */
 };
 
 /* typedef DIR - not the same as Unix */
+struct DIR_W32 {
+	HANDLE handle;			/* _findfirst/_findnext handle */
+	uint16_t offset;		/* offset into directory */
+	uint8_t finished;		/* 1 if there are not more files */
+	WIN32_FIND_DATAW fileinfo;	/* from _findfirst/_findnext */
+	wchar_t *dirw;			/* the dir we are reading */
+	struct dirent dent;		/* the dirent to return */
+};
 typedef struct DIR_W32 DIR;
 
 /* Function prototypes */