shell: Always use explicit large file API

Commit Message

Herbert Xu May 7, 2020, 1:42 p.m. UTC

There are some remaining stat/readdir calls in dash that may lead
to spurious EOVERFLOW errors on 32-bit platforms.  This patch changes
them (as well as open(2)) to use the explicit large file API.

Reported-by: Tatsuki Sugiura <[e-mail]>
Signed-off-by: Herbert Xu <[e-mail]>

Comments

Jilles Tjoelker May 9, 2020, 1:13 p.m. UTC | #1

On Thu, May 07, 2020 at 11:42:12PM +1000, Herbert Xu wrote:
> There are some remaining stat/readdir calls in dash that may lead
> to spurious EOVERFLOW errors on 32-bit platforms.  This patch changes
> them (as well as open(2)) to use the explicit large file API.

> Reported-by: Tatsuki Sugiura <[e-mail]>
> Signed-off-by: Herbert Xu <[e-mail]>

> diff --git a/configure.ac b/configure.ac
> index 5dab5aa..dbd97d8 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -144,8 +144,13 @@ AC_CHECK_FUNC(stat64,, [
>  	AC_DEFINE(fstat64, fstat, [64-bit operations are the same as 32-bit])
>  	AC_DEFINE(lstat64, lstat, [64-bit operations are the same as 32-bit])
>  	AC_DEFINE(stat64, stat, [64-bit operations are the same as 32-bit])
> +	AC_DEFINE(readdir64, readdir,
> +		  [64-bit operations are the same as 32-bit])
> +	AC_DEFINE(dirent64, dirent,
> +		  [64-bit operations are the same as 32-bit])
>  ])
> [snip]

Is it possible to use AC_SYS_LARGEFILE and the normal
fstat/lstat/readdir/dirent/open/etc instead of the non-standard "64"
interfaces?

I understand that dash formerly used the "64" interfaces selectively, so
that the binary could be a bit smaller by avoiding 64-bit numbers where
they were not necessary. Now that the feature "supports files with inode
numbers that do not fit in 32 bits" is considered essential, this
complexity seems unnecessary.

I'm sorry for not providing a patch.

Patch

diff --git a/configure.ac b/configure.ac
index 5dab5aa..dbd97d8 100644
--- a/configure.ac
+++ b/configure.ac
@@ -144,8 +144,13 @@  AC_CHECK_FUNC(stat64,, [
 	AC_DEFINE(fstat64, fstat, [64-bit operations are the same as 32-bit])
 	AC_DEFINE(lstat64, lstat, [64-bit operations are the same as 32-bit])
 	AC_DEFINE(stat64, stat, [64-bit operations are the same as 32-bit])
+	AC_DEFINE(readdir64, readdir,
+		  [64-bit operations are the same as 32-bit])
+	AC_DEFINE(dirent64, dirent,
+		  [64-bit operations are the same as 32-bit])
 ])
 
+dnl OS X apparently has stat64 but not open64.
 AC_CHECK_FUNC(open64,, [
 	AC_DEFINE(open64, open, [64-bit operations are the same as 32-bit])
 ])
diff --git a/src/bltin/test.c b/src/bltin/test.c
index b7188df..c7fc479 100644
--- a/src/bltin/test.c
+++ b/src/bltin/test.c
@@ -473,17 +473,17 @@  static int isoperand(char **tp)
 static int
 newerf (const char *f1, const char *f2)
 {
-	struct stat b1, b2;
+	struct stat64 b1, b2;
 
 #ifdef HAVE_ST_MTIM
-	return (stat (f1, &b1) == 0 &&
-		stat (f2, &b2) == 0 &&
+	return (stat64(f1, &b1) == 0 &&
+		stat64(f2, &b2) == 0 &&
 		( b1.st_mtim.tv_sec > b2.st_mtim.tv_sec ||
 		 (b1.st_mtim.tv_sec == b2.st_mtim.tv_sec && (b1.st_mtim.tv_nsec > b2.st_mtim.tv_nsec )))
 	);
 #else
-	return (stat (f1, &b1) == 0 &&
-		stat (f2, &b2) == 0 &&
+	return (stat64(f1, &b1) == 0 &&
+		stat64(f2, &b2) == 0 &&
 		b1.st_mtime > b2.st_mtime);
 #endif
 }
@@ -491,17 +491,17 @@  newerf (const char *f1, const char *f2)
 static int
 olderf (const char *f1, const char *f2)
 {
-	struct stat b1, b2;
+	struct stat64 b1, b2;
 
 #ifdef HAVE_ST_MTIM
-	return (stat (f1, &b1) == 0 &&
-		stat (f2, &b2) == 0 &&
+	return (stat64(f1, &b1) == 0 &&
+		stat64(f2, &b2) == 0 &&
 		(b1.st_mtim.tv_sec < b2.st_mtim.tv_sec ||
 		 (b1.st_mtim.tv_sec == b2.st_mtim.tv_sec && (b1.st_mtim.tv_nsec < b2.st_mtim.tv_nsec )))
 	);
 #else
-	return (stat (f1, &b1) == 0 &&
-		stat (f2, &b2) == 0 &&
+	return (stat64(f1, &b1) == 0 &&
+		stat64(f2, &b2) == 0 &&
 		b1.st_mtime < b2.st_mtime);
 #endif
 }
@@ -509,10 +509,10 @@  olderf (const char *f1, const char *f2)
 static int
 equalf (const char *f1, const char *f2)
 {
-	struct stat b1, b2;
+	struct stat64 b1, b2;
 
-	return (stat (f1, &b1) == 0 &&
-		stat (f2, &b2) == 0 &&
+	return (stat64(f1, &b1) == 0 &&
+		stat64(f2, &b2) == 0 &&
 		b1.st_dev == b2.st_dev &&
 		b1.st_ino == b2.st_ino);
 }
diff --git a/src/cd.c b/src/cd.c
index b6742af..1ef1dc5 100644
--- a/src/cd.c
+++ b/src/cd.c
@@ -96,7 +96,7 @@  cdcmd(int argc, char **argv)
 	const char *path;
 	const char *p;
 	char c;
-	struct stat statb;
+	struct stat64 statb;
 	int flags;
 	int len;
 
@@ -132,7 +132,7 @@  dotdot:
 		c = *p;
 		p = stalloc(len);
 
-		if (stat(p, &statb) >= 0 && S_ISDIR(statb.st_mode)) {
+		if (stat64(p, &statb) >= 0 && S_ISDIR(statb.st_mode)) {
 			if (c && c != ':')
 				flags |= CD_PRINT;
 docd:
diff --git a/src/expand.c b/src/expand.c
index 4a5d75a..ecd7ee5 100644
--- a/src/expand.c
+++ b/src/expand.c
@@ -1286,7 +1286,7 @@  expmeta(char *name, unsigned name_len, unsigned expdir_len)
 	int metaflag;
 	struct stat64 statb;
 	DIR *dirp;
-	struct dirent *dp;
+	struct dirent64 *dp;
 	int atend;
 	int matchdot;
 	int esc;
@@ -1363,7 +1363,7 @@  expmeta(char *name, unsigned name_len, unsigned expdir_len)
 		p++;
 	if (*p == '.')
 		matchdot++;
-	while (! int_pending() && (dp = readdir(dirp)) != NULL) {
+	while (! int_pending() && (dp = readdir64(dirp)) != NULL) {
 		if (dp->d_name[0] == '.' && ! matchdot)
 			continue;
 		if (pmatch(start, dp->d_name)) {
diff --git a/src/input.c b/src/input.c
index 177fd0a..7d6be63 100644
--- a/src/input.c
+++ b/src/input.c
@@ -397,7 +397,7 @@  setinputfile(const char *fname, int flags)
 	int fd;
 
 	INTOFF;
-	if ((fd = open(fname, O_RDONLY)) < 0) {
+	if ((fd = open64(fname, O_RDONLY)) < 0) {
 		if (flags & INPUT_NOFILE_OK)
 			goto out;
 		exitstatus = 127;
diff --git a/src/jobs.c b/src/jobs.c
index a9e6524..f65435d 100644
--- a/src/jobs.c
+++ b/src/jobs.c
@@ -196,7 +196,7 @@  setjobctl(int on)
 		return;
 	if (on) {
 		int ofd;
-		ofd = fd = open(_PATH_TTY, O_RDWR);
+		ofd = fd = open64(_PATH_TTY, O_RDWR);
 		if (fd < 0) {
 			fd += 3;
 			while (!isatty(fd))
@@ -887,7 +887,7 @@  static void forkchild(struct job *jp, union node *n, int mode)
 		ignoresig(SIGQUIT);
 		if (jp->nprocs == 0) {
 			close(0);
-			if (open(_PATH_DEVNULL, O_RDONLY) != 0)
+			if (open64(_PATH_DEVNULL, O_RDONLY) != 0)
 				sh_error("Can't open %s", _PATH_DEVNULL);
 		}
 	}
diff --git a/src/main.c b/src/main.c
index 36431fc..7a28534 100644
--- a/src/main.c
+++ b/src/main.c
@@ -298,7 +298,7 @@  find_dot_file(char *basename)
 {
 	char *fullname;
 	const char *path = pathval();
-	struct stat statb;
+	struct stat64 statb;
 	int len;
 
 	/* don't try this for absolute or relative paths */
@@ -308,7 +308,7 @@  find_dot_file(char *basename)
 	while ((len = padvance(&path, basename)) >= 0) {
 		fullname = stackblock();
 		if ((!pathopt || *pathopt == 'f') &&
-		    !stat(fullname, &statb) && S_ISREG(statb.st_mode)) {
+		    !stat64(fullname, &statb) && S_ISREG(statb.st_mode)) {
 			/* This will be freed by the caller. */
 			return stalloc(len);
 		}
diff --git a/src/var.c b/src/var.c
index b146018..ef9c2bd 100644
--- a/src/var.c
+++ b/src/var.c
@@ -125,7 +125,7 @@  INIT {
 	char **envp;
 	static char ppid[32] = "PPID=";
 	const char *p;
-	struct stat st1, st2;
+	struct stat64 st1, st2;
 
 	initvar();
 	for (envp = environ ; *envp ; envp++) {
@@ -143,7 +143,7 @@  INIT {
 
 	p = lookupvar("PWD");
 	if (p)
-		if (*p != '/' || stat(p, &st1) || stat(".", &st2) ||
+		if (*p != '/' || stat64(p, &st1) || stat64(".", &st2) ||
 		    st1.st_dev != st2.st_dev || st1.st_ino != st2.st_ino)
 			p = 0;
 	setpwd(p, 0);