[DFSAN] Add support for strncat

[tkuchta]

Summary

This patch adds a support for the libc strncat() function in DFSAN

[browneee]

compiler-rt/test/dfsan/custom.cpp

377

Should we test the case where n < strlen(src) ?

In this case, the taint labels on the later bytes (not overridden due to the limit) should be unchanged, in addition to the taint labels on the middle changed bytes being changed.

I'd probably duplicate the function to do this.

[tkuchta]

Thank you for the comments. I have prepared and updated version with tests accounting for n < strlen(src)

[browneee]

compiler-rt/test/dfsan/custom.cpp

342–348

Change back to

char *ret = strcat(p, src);

[tkuchta]

Right, my bad - updated

[tkuchta]

Hello, please let me know if there is anything more to fix.

compiler-rt/lib/dfsan/dfsan_custom.cpp

SANITIZER_INTERFACE_ATTRIBUTE char *__dfso_strcat(
   size_t src_len = strlen(src);
   dfsan_mem_origin_transfer(dest + dest_len, src, src_len);
   dfsan_mem_shadow_transfer(dest + dest_len, src, src_len);
   *ret_label = dest_label;
   *ret_origin = dest_origin;
   return ret;
 }
 
+SANITIZER_INTERFACE_ATTRIBUTE char *__dfsw_strncat(
+    char *dest, const char *src, size_t num, dfsan_label dest_label,
+    dfsan_label src_label, dfsan_label num_label, dfsan_label *ret_label) {
+  size_t src_len = strlen(src);
+  src_len = src_len < num ? src_len : num;
+  size_t dest_len = strlen(dest);
+
+  char *ret = strncat(dest, src, num);
+  dfsan_mem_shadow_transfer(dest + dest_len, src, src_len);
+  *ret_label = dest_label;
+  return ret;
+}
+
+SANITIZER_INTERFACE_ATTRIBUTE char *__dfso_strncat(
+    char *dest, const char *src, size_t num, dfsan_label dest_label,
+    dfsan_label src_label, dfsan_label num_label, dfsan_label *ret_label,
+    dfsan_origin dest_origin, dfsan_origin src_origin, dfsan_origin num_origin,
+    dfsan_origin *ret_origin) {
+  size_t src_len = strlen(src);
+  src_len = src_len < num ? src_len : num;
+  size_t dest_len = strlen(dest);
+
+  char *ret = strncat(dest, src, num);
+
+  dfsan_mem_origin_transfer(dest + dest_len, src, src_len);
+  dfsan_mem_shadow_transfer(dest + dest_len, src, src_len);
+  *ret_label = dest_label;
+  *ret_origin = dest_origin;
+  return ret;
+}
+
 SANITIZER_INTERFACE_ATTRIBUTE char *
 __dfsw_strdup(const char *s, dfsan_label s_label, dfsan_label *ret_label) {
   size_t len = strlen(s);
   void *p = malloc(len+1);
   dfsan_memcpy(p, s, len+1);
   *ret_label = 0;
   return static_cast<char *>(p);
 }

compiler-rt/lib/dfsan/done_abilist.txt

 fun:isalpha=functional
 fun:isdigit=functional
 fun:isprint=functional
 fun:isxdigit=functional
 fun:isalnum=functional
 fun:ispunct=functional
 fun:isspace=functional
 fun:tolower=functional
+fun:_tolower=functional
 fun:toupper=functional
 
 # Functions that return a value that is data-dependent on the input.
 fun:__isinf=functional
 fun:__isinff=functional
 fun:__signbit=functional
 fun:__signbitf=functional
 fun:__signbitl=functional

 fun:strdup=custom
 fun:strncpy=custom
 fun:strtod=custom
 fun:strtol=custom
 fun:strtoll=custom
 fun:strtoul=custom
 fun:strtoull=custom
 fun:strcat=custom
+fun:strncat=custom
 
 # Functions that produce an output that is computed from the input, but is not
 # necessarily data dependent.
 fun:bcmp=custom
 fun:memchr=custom
 fun:memcmp=custom
 fun:strcasecmp=custom
 fun:strchr=custom

compiler-rt/test/dfsan/custom.cpp

#ifdef STRICT_DATA_DEPENDENCIES
   ASSERT_ZERO_LABEL(rv);
   ASSERT_ZERO_ORIGIN(rv);
 #else
   ASSERT_LABEL(rv, i_label);
   ASSERT_EQ_ORIGIN(rv, str1[3]);
 #endif
 }
 
-void test_strcat() {
+void _test_strcat(int _strncat) {
   char src[] = "world";
   int volatile x = 0; // buffer to ensure src and dst do not share origins
   (void)x;
   char dst[] = "hello \0    ";
   int volatile y = 0; // buffer to ensure dst and p do not share origins
   (void)y;
   char *p = dst;
   dfsan_set_label(k_label, &p, sizeof(p));
   dfsan_set_label(i_label, src, sizeof(src));
   dfsan_set_label(j_label, dst, sizeof(dst));
   dfsan_origin dst_o = dfsan_get_origin((long)dst[0]);
   (void)dst_o;
-  char *ret = strcat(p, src);
+  char *ret = 0;
+  if (!_strncat)
+    ret = strcat(p, src);
+  else
+    ret = strncat(p, src, strlen(src));
+
   ASSERT_LABEL(ret, k_label);
   ASSERT_EQ_ORIGIN(ret, p);
   assert(ret == dst);
   assert(strcmp(src, dst + 6) == 0);
   // Origins are assigned for every 4 contiguous 4-aligned bytes. After
   // appending src to dst, origins of src can overwrite origins of dst if their
   // application adddresses are within [start_aligned_down, end_aligned_up).
   // Other origins are not changed.
   char *start_aligned_down = (char *)(((size_t)(dst + 6)) & ~3UL);
void _test_strcat(int _strncat) {
   }
   for (int i = 6; i < strlen(dst); ++i) {
     ASSERT_LABEL(dst[i], i_label);
     assert(dfsan_get_label(dst[i]) == dfsan_get_label(src[i - 6]));
   }
   ASSERT_LABEL(dst[11], j_label);
 }
 
+void test_strcat() { _test_strcat(0); }
+
+void test_strncat() { _test_strcat(1); }
+
 void test_strlen() {
   char str1[] = "str1";
   dfsan_set_label(i_label, &str1[3], 1);
 
   int rv = strlen(str1);
   assert(rv == 4);
 #ifdef STRICT_DATA_DEPENDENCIES
   ASSERT_ZERO_LABEL(rv);
int main(void) {
   test_snprintf();
   test_socketpair();
   test_sprintf();
   test_stat();
   test_strcasecmp();
   test_strchr();
   test_strcmp();
   test_strcat();
+  test_strncat();
   test_strcpy();
   test_strdup();
   test_strlen();
   test_strncasecmp();
   test_strncmp();
   test_strncpy();
   test_strpbrk();
   test_strsep();