Implement a custom print formatter (archive_string_vsprintf)

for libarchive error messages.  Mostly, this
avoids a portability headache related to
copying va_list arguments (some FreeBSD 5
platforms require va_copy; FreeBSD 4 doesn't
support va_copy at all).  It also dramatically reduces the
size of libarchive for embedded applications:
a minimal "untar" program using libarchive can now be
under 64k statically linked (as opposed to ~100k
using library *printf() functions).

MFC after: 14 days

4 files changed, 90 insertions, 14 deletions

diff --git a/lib/libarchive/archive_platform.h b/lib/libarchive/archive_platform.h
index 490bcba..0e98ad3 100644
--- a/lib/libarchive/archive_platform.h
+++ b/lib/libarchive/archive_platform.h
@@ -105,6 +105,12 @@
 #include <inttypes.h>
 #endif
 
+/* FreeBSD 4 and earlier lack intmax_t/uintmax_t */
+#if defined(__FreeBSD__) && __FreeBSD__ < 5
+#define	intmax_t int64_t
+#define	uintmax_t uint64_t
+#endif
+
 /* TODO: Test for the functions we use as well... */
 #if HAVE_SYS_ACL_H
 #define	HAVE_POSIX_ACLS	1
diff --git a/lib/libarchive/archive_string.h b/lib/libarchive/archive_string.h
index f4e8b76..c9d1c00 100644
--- a/lib/libarchive/archive_string.h
+++ b/lib/libarchive/archive_string.h
@@ -67,6 +67,7 @@ __archive_strappend_char_UTF8(struct archive_string *, int);
 /* Append an integer in the specified base (2 <= base <= 16). */
 struct archive_string *
 __archive_strappend_int(struct archive_string *as, int d, int base);
+#define	archive_strappend_int __archive_strappend_int
 
 /* Basic append operation. */
 struct archive_string *
diff --git a/lib/libarchive/archive_string_sprintf.c b/lib/libarchive/archive_string_sprintf.c
index d5e8d86..64ca7d0 100644
--- a/lib/libarchive/archive_string_sprintf.c
+++ b/lib/libarchive/archive_string_sprintf.c
@@ -28,10 +28,16 @@
 __FBSDID("$FreeBSD$");
 
 /*
- * This uses 'printf' family functions, which can cause issues
- * for size-critical applications.  I've separated it out to make
- * this issue clear.  (Currently, it is called directly from within
- * the core code, so it cannot easily be omitted.)
+ * The use of printf()-family functions can be troublesome
+ * for space-constrained applications.  In addition, correctly
+ * implementing this function in terms of vsnprintf() requires
+ * two calls (one to determine the size, another to format the
+ * result), which in turn requires duplicating the argument list
+ * using va_copy, which isn't yet universally available.
+ *
+ * So, I've implemented a bare minimum of printf()-like capability
+ * here.  This is only used to format error messages, so doesn't
+ * require any floating-point support or field-width handling.
  */
 
 #include <stdio.h>
@@ -46,21 +52,78 @@ void
 __archive_string_vsprintf(struct archive_string *as, const char *fmt,
     va_list ap)
 {
-	size_t l;
-	va_list ap1;
+	char long_flag;
+	intmax_t s; /* Signed integer temp. */
+	uintmax_t u; /* Unsigned integer temp. */
+	const char *p, *p2;
+
+	__archive_string_ensure(as, 64);
 
 	if (fmt == NULL) {
 		as->s[0] = 0;
 		return;
 	}
 
-	va_copy(ap1, ap);
-	l = vsnprintf(as->s, as->buffer_length, fmt, ap);
-	/* If output is bigger than the buffer, resize and try again. */
-	if (l+1 >= as->buffer_length) {
-		__archive_string_ensure(as, l + 1);
-		l = vsnprintf(as->s, as->buffer_length, fmt, ap1);
+	long_flag = '\0';
+	for (p = fmt; *p != '\0'; p++) {
+		const char *saved_p = p;
+
+		if (*p != '%') {
+			archive_strappend_char(as, *p);
+			continue;
+		}
+
+		p++;
+
+		switch(*p) {
+		case 'j':
+			long_flag = 'j';
+			p++;
+			break;
+		case 'l':
+			long_flag = 'l';
+			p++;
+			break;
+		}
+
+		switch (*p) {
+		case '%':
+			__archive_strappend_char(as, '%');
+			break;
+		case 'c':
+			s = va_arg(ap, int);
+			__archive_strappend_char(as, s);
+			break;
+		case 'd':
+			switch(long_flag) {
+			case 'j': s = va_arg(ap, intmax_t); break;
+			case 'l': s = va_arg(ap, long); break;
+			default:  s = va_arg(ap, int); break;
+			}
+			archive_strappend_int(as, s, 10);
+			break;
+		case 's':
+			p2 = va_arg(ap, char *);
+			archive_strcat(as, p2);
+			break;
+		case 'o': case 'u': case 'x': case 'X':
+			/* Common handling for unsigned integer formats. */
+			switch(long_flag) {
+			case 'j': u = va_arg(ap, uintmax_t); break;
+			case 'l': u = va_arg(ap, unsigned long); break;
+			default:  u = va_arg(ap, unsigned int); break;
+			}
+			/* Format it in the correct base. */
+			switch (*p) {
+			case 'o': archive_strappend_int(as, u, 8); break;
+			case 'u': archive_strappend_int(as, u, 10); break;
+			default: archive_strappend_int(as, u, 16); break;
+			}
+			break;
+		default:
+			/* Rewind and print the initial '%' literally. */
+			p = saved_p;
+			archive_strappend_char(as, *p);
+		}
 	}
-	as->length = l;
-	va_end(ap1);
 }
diff --git a/lib/libarchive/configure.ac.in b/lib/libarchive/configure.ac.in
index 2bb657c..0068c8d 100644
--- a/lib/libarchive/configure.ac.in
+++ b/lib/libarchive/configure.ac.in
@@ -36,6 +36,9 @@ AC_CHECK_TYPE(__int64_t, [long long])
 AC_CHECK_TYPE(_int64_t, [__int64_t])
 AC_CHECK_TYPE(int64_t, [_int64_t])
 
+# Likewise, consider defining "intmax_t" in terms of int64_t
+AC_CHECK_TYPE(intmax_t, [int64_t])
+
 #
 # If any of the common 64-bit unsigned types is defined, set "uint64_t"
 #
@@ -44,6 +47,9 @@ AC_CHECK_TYPE(_uint64_t, [__uint64_t])
 AC_CHECK_TYPE(u_int64_t, [_uint64_t])
 AC_CHECK_TYPE(uint64_t, [u_int64_t])
 
+# Likewise, consider defining "uintmax_t" in terms of uint64_t
+AC_CHECK_TYPE(uintmax_t, [uint64_t])
+
 AC_CHECK_MEMBERS([struct stat.st_rdev, struct stat.st_mtimespec.tv_nsec, struct stat.st_mtim.tv_nsec])
 
 AC_CHECK_DECL([EFTYPE],