From 377709664480a30fa5acdd11c7ca8c16669678ce Mon Sep 17 00:00:00 2001
From: Tavian Barnes <tavianator@tavianator.com>
Date: Wed, 6 Sep 2023 14:59:59 -0400
Subject: bfstd: Fix an OOB string index in xmbrtowc()

This bug could be reproduced with something like

    $ bfs -samefile $'\xFA\xFA'
    bfs: error: bfs: dstrnescat@src/dstring.c:252: wordesc() result truncated

or worse, with -DNDEBUG,

    $ bfs -samefile $'.....................\xFA\xFA'
    bfs: error: bfs -samefile $'.....................\xFA\xFA\x00\x55\x53\x45\x52\x3D\x74\x61\x76\x69\x61\x6E\x61\x74\x6F\x72
    bfs: error:               ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
    bfs: error: No such file or directory.

which prints the memory after the end of the string (in this case, the
environment variable USER=tavianator).

The bug was caused by the line `*i += len`, which was intended to be
`*i = len`.  But actually, the right behaviour seems to be `*i += 1`.

Fixes: 19c96abe0a1ee56cf206fd5e87defb1fd3e0daa5
---
 src/bfstd.c                      |  8 ++------
 tests/bfstd.c                    | 18 ++++++++++++++++++
 tests/common/samefile_wordesc.sh |  4 ++++
 3 files changed, 24 insertions(+), 6 deletions(-)
 create mode 100644 tests/common/samefile_wordesc.sh

diff --git a/src/bfstd.c b/src/bfstd.c
index e546a47..2d9f60a 100644
--- a/src/bfstd.c
+++ b/src/bfstd.c
@@ -546,15 +546,11 @@ int xstrtofflags(const char **str, unsigned long long *set, unsigned long long *
 static int xmbrtowc(wchar_t *wc, size_t *i, const char *str, size_t len, mbstate_t *mb) {
 	size_t mblen = mbrtowc(wc, str + *i, len - *i, mb);
 	switch (mblen) {
-	case -1:
-		// Invalid byte sequence
+	case -1: // Invalid byte sequence
+	case -2: // Incomplete byte sequence
 		*i += 1;
 		memset(mb, 0, sizeof(*mb));
 		return -1;
-	case -2:
-		// Incomplete byte sequence
-		*i += len;
-		return -1;
 	default:
 		*i += mblen;
 		return 0;
diff --git a/tests/bfstd.c b/tests/bfstd.c
index fa854a8..2db084a 100644
--- a/tests/bfstd.c
+++ b/tests/bfstd.c
@@ -23,6 +23,15 @@ static void check_base_dir(const char *path, const char *dir, const char *base)
 	free(xbase);
 }
 
+/** Check the result of wordesc(). */
+static void check_wordesc(const char *str, const char *exp, enum wesc_flags flags) {
+	char buf[256];
+	char *end = buf + sizeof(buf);
+	char *ret = wordesc(buf, end, str, flags);
+	bfs_verify(ret != end);
+	bfs_verify(strcmp(buf, exp) == 0, "wordesc(%s) == %s (!= %s)", str, buf, exp);
+}
+
 int main(void) {
 	// From man 3p basename
 	check_base_dir("usr", ".", "usr");
@@ -36,5 +45,14 @@ int main(void) {
 	check_base_dir("//usr//lib//", "//usr", "lib");
 	check_base_dir("/home//dwc//test", "/home//dwc", "test");
 
+	check_wordesc("", "\"\"", WESC_SHELL);
+	check_wordesc("word", "word", WESC_SHELL);
+	check_wordesc("two words", "\"two words\"", WESC_SHELL);
+	check_wordesc("word's", "\"word's\"", WESC_SHELL);
+	check_wordesc("\"word\"", "'\"word\"'", WESC_SHELL);
+	check_wordesc("\"word's\"", "'\"word'\\''s\"'", WESC_SHELL);
+	check_wordesc("\033[1mbold's\033[0m", "$'\\e[1mbold\\'s\\e[0m'", WESC_SHELL | WESC_TTY);
+	check_wordesc("\x7F", "$'\\x7F'", WESC_SHELL | WESC_TTY);
+
 	return EXIT_SUCCESS;
 }
diff --git a/tests/common/samefile_wordesc.sh b/tests/common/samefile_wordesc.sh
new file mode 100644
index 0000000..b5d158f
--- /dev/null
+++ b/tests/common/samefile_wordesc.sh
@@ -0,0 +1,4 @@
+# Regression test: don't abort on incomplete UTF-8 sequences
+export LC_ALL=$(locale -a | grep -Ei 'utf-?8$' | head -n1)
+test -n "$LC_ALL" || skip
+! invoke_bfs -samefile $'\xFA\xFA'
-- 
cgit v1.2.3