libarchive: merge bugfixes from vendor branch

 CTSRD-CHERI#2147 archive_string: clean up strncat_from_utf8_to_utf8 (36047967a)
 CTSRD-CHERI#2153 archive_match: check archive_read_support_format_raw()
       return value (0ce1b4c38)
 CTSRD-CHERI#2154 archive_match: turn counter into flag (287e05d53)
 CTSRD-CHERI#2155 lha: Do not allow negative file sizes (93b11caed)
 CTSRD-CHERI#2156 tests: setenv LANG to en_US.UTF-8 in bsdunzip test_I.c (83e8b0ea8)

MFC after:	3 days

Author: mmatuska

contrib/libarchive/libarchive/archive_match.c

@@ -46,7 +46,7 @@
 
 struct match {
 	struct match		*next;
-	int			 matches;
+	int			 matched;
 	struct archive_mstring	 pattern;
 };
 
@@ -605,7 +605,8 @@ add_pattern_from_file(struct archive_match *a, struct match_list *mlist,
 		return (ARCHIVE_FATAL);
 	}
 	r = archive_read_support_format_raw(ar);
-	r = archive_read_support_format_empty(ar);
+	if (r == ARCHIVE_OK)
+		r = archive_read_support_format_empty(ar);
 	if (r != ARCHIVE_OK) {
 		archive_copy_error(&(a->archive), ar);
 		archive_read_free(ar);
@@ -724,12 +725,12 @@ path_excluded(struct archive_match *a, int mbs, const void *pathname)
 	matched = NULL;
 	for (match = a->inclusions.first; match != NULL;
 	    match = match->next){
-		if (match->matches == 0 &&
+		if (!match->matched &&
 		    (r = match_path_inclusion(a, match, mbs, pathname)) != 0) {
 			if (r < 0)
 				return (r);
 			a->inclusions.unmatched_count--;
-			match->matches++;
+			match->matched = 1;
 			matched = match;
 		}
 	}
@@ -752,11 +753,10 @@ path_excluded(struct archive_match *a, int mbs, const void *pathname)
 	for (match = a->inclusions.first; match != NULL;
 	    match = match->next){
 		/* We looked at previously-unmatched inclusions already. */
-		if (match->matches > 0 &&
+		if (match->matched &&
 		    (r = match_path_inclusion(a, match, mbs, pathname)) != 0) {
 			if (r < 0)
 				return (r);
-			match->matches++;
 			return (0);
 		}
 	}
@@ -879,7 +879,7 @@ match_list_unmatched_inclusions_next(struct archive_match *a,
 	for (m = list->unmatched_next; m != NULL; m = m->next) {
 		int r;
 
-		if (m->matches)
+		if (m->matched)
 			continue;
 		if (mbs) {
 			const char *p;
@@ -1793,7 +1793,7 @@ match_owner_name_mbs(struct archive_match *a, struct match_list *list,
 		    < 0 && errno == ENOMEM)
 			return (error_nomem(a));
 		if (p != NULL && strcmp(p, name) == 0) {
-			m->matches++;
+			m->matched = 1;
 			return (1);
 		}
 	}
@@ -1814,7 +1814,7 @@ match_owner_name_wcs(struct archive_match *a, struct match_list *list,
 		    < 0 && errno == ENOMEM)
 			return (error_nomem(a));
 		if (p != NULL && wcscmp(p, name) == 0) {
-			m->matches++;
+			m->matched = 1;
 			return (1);
 		}
 	}

contrib/libarchive/libarchive/archive_read_support_format_lha.c

@@ -1347,6 +1347,8 @@ lha_read_file_extended_header(struct archive_read *a, struct lha *lha,
 				lha->compsize = archive_le64dec(extdheader);
 				extdheader += sizeof(uint64_t);
 				lha->origsize = archive_le64dec(extdheader);
+				if (lha->compsize < 0 || lha->origsize < 0)
+					goto invalid;
 			}
 			break;
 		case EXT_CODEPAGE:

contrib/libarchive/libarchive/archive_string.c

@@ -2640,81 +2640,69 @@ unicode_to_utf16le(char *p, size_t remaining, uint32_t uc)
 }
 
 /*
- * Copy UTF-8 string in checking surrogate pair.
- * If any surrogate pair are found, it would be canonicalized.
+ * Append new UTF-8 string to existing UTF-8 string.
+ * Existing string is assumed to already be in proper form;
+ * the new string will have invalid sequences replaced and
+ * surrogate pairs canonicalized.
  */
 static int
-strncat_from_utf8_to_utf8(struct archive_string *as, const void *_p,
+strncat_from_utf8_to_utf8(struct archive_string *as, const void *_src,
     size_t len, struct archive_string_conv *sc)
 {
-	const char *s;
-	char *p, *endp;
-	int n, ret = 0;
-
+	int ret = 0;
+	const char *src = _src;
 	(void)sc; /* UNUSED */
 
+	/* Pre-extend the destination */
 	if (archive_string_ensure(as, as->length + len + 1) == NULL)
 		return (-1);
 
-	s = (const char *)_p;
-	p = as->s + as->length;
-	endp = as->s + as->buffer_length -1;
-	do {
+	/* Invariant: src points to the first UTF8 byte that hasn't
+	 * been copied to the destination `as`. */
+	for (;;) {
+		int n;
 		uint32_t uc;
-		const char *ss = s;
-		size_t w;
+		const char *e = src;
 
-		/*
-		 * Forward byte sequence until a conversion of that is needed.
-		 */
-		while ((n = utf8_to_unicode(&uc, s, len)) > 0) {
-			s += n;
+		/* Skip UTF-8 sequences until we reach end-of-string or
+		 * a code point that needs conversion. */
+		while ((n = utf8_to_unicode(&uc, e, len)) > 0) {
+			e += n;
 			len -= n;
 		}
-		if (ss < s) {
-			if (p + (s - ss) > endp) {
-				as->length = p - as->s;
-				if (archive_string_ensure(as,
-				    as->buffer_length + len + 1) == NULL)
-					return (-1);
-				p = as->s + as->length;
-				endp = as->s + as->buffer_length -1;
-			}
-
-			memcpy(p, ss, s - ss);
-			p += s - ss;
+		/* Copy the part that doesn't need conversion */
+		if (e > src) {
+			if (archive_string_append(as, src, e - src) == NULL)
+				return (-1);
+			src = e;
 		}
 
-		/*
-		 * If n is negative, current byte sequence needs a replacement.
-		 */
-		if (n < 0) {
+		if (n == 0) {
+			/* We reached end-of-string */
+			return (ret);
+		} else {
+			/* Next code point needs conversion */
+			char t[4];
+			size_t w;
+
+			/* Try decoding a surrogate pair */
 			if (n == -3 && IS_SURROGATE_PAIR_LA(uc)) {
-				/* Current byte sequence may be CESU-8. */
-				n = cesu8_to_unicode(&uc, s, len);
+				n = cesu8_to_unicode(&uc, src, len);
 			}
+			/* Not a (valid) surrogate, so use a replacement char */
 			if (n < 0) {
-				ret = -1;
-				n *= -1;/* Use a replaced unicode character. */
-			}
-
-			/* Rebuild UTF-8 byte sequence. */
-			while ((w = unicode_to_utf8(p, endp - p, uc)) == 0) {
-				as->length = p - as->s;
-				if (archive_string_ensure(as,
-				    as->buffer_length + len + 1) == NULL)
-					return (-1);
-				p = as->s + as->length;
-				endp = as->s + as->buffer_length -1;
+				ret = -1; /* Return -1 if we used any replacement */
+				n *= -1;
 			}
-			p += w;
-			s += n;
+			/* Consume converted code point */
+			src += n;
 			len -= n;
+			/* Convert and append new UTF-8 sequence. */
+			w = unicode_to_utf8(t, sizeof(t), uc);
+			if (archive_string_append(as, t, w) == NULL)
+				return (-1);
 		}
-	} while (n > 0);
-	as->length = p - as->s;
-	as->s[as->length] = '\0';
-	return (ret);
+	}
 }
 
 static int

contrib/libarchive/unzip/test/test_I.c

@@ -33,6 +33,7 @@
 DEFINE_TEST(test_I)
 {
 	const char *reffile = "test_I.zip";
+	const char *lang;
 	int r;
 
 #if HAVE_SETLOCALE
@@ -44,11 +45,18 @@ DEFINE_TEST(test_I)
 	skipping("setlocale() not available on this system.");
 #endif
 
+	lang = getenv("LANG");
+	setenv("LANG", "en_US.UTF-8", 1);
 	extract_reference_file(reffile);
 	r = systemf("%s -I UTF-8 %s >test.out 2>test.err", testprog, reffile);
 	assertEqualInt(0, r);
 	assertNonEmptyFile("test.out");
 	assertEmptyFile("test.err");
 
 	assertTextFileContents("Hello, World!\n", "Γειά σου Κόσμε.txt");
+
+	if (lang == NULL)
+		unsetenv("LANG");
+	else
+		setenv("LANG", lang, 1);
 }