[compiler-rt] 1601879 - [compiler-rt][nsan] Fix strsep interceptor (#106307)

[via llvm-commits]

Author: Alexander Shaposhnikov
Date: 2024-08-27T17:40:13-07:00
New Revision: 1601879f5d690595889a5537c0049b62df4657c7

URL: https://github.com/llvm/llvm-project/commit/1601879f5d690595889a5537c0049b62df4657c7
DIFF: https://github.com/llvm/llvm-project/commit/1601879f5d690595889a5537c0049b62df4657c7.diff

LOG: [compiler-rt][nsan] Fix strsep interceptor (#106307)

Fix strsep interceptor.
For strsep description see
https://www.man7.org/linux/man-pages/man3/strsep.3.html

Added: 
    compiler-rt/test/nsan/intercep_strsep.cpp

Modified: 
    compiler-rt/lib/nsan/nsan_interceptors.cpp

Removed: 
    


################################################################################
diff  --git a/compiler-rt/lib/nsan/nsan_interceptors.cpp b/compiler-rt/lib/nsan/nsan_interceptors.cpp
index d607c8d6a636b8..5f25eba8560705 100644
--- a/compiler-rt/lib/nsan/nsan_interceptors.cpp
+++ b/compiler-rt/lib/nsan/nsan_interceptors.cpp
@@ -94,7 +94,7 @@ INTERCEPTOR(char *, strfry, char *s) {
 
 INTERCEPTOR(char *, strsep, char **Stringp, const char *delim) {
   char *OrigStringp = REAL(strsep)(Stringp, delim);
-  if (Stringp != nullptr) {
+  if (*Stringp != nullptr) {
     // The previous character has been overwritten with a '\0' char.
     __nsan_set_value_unknown(reinterpret_cast<u8 *>(*Stringp) - 1, 1);
   }

diff  --git a/compiler-rt/test/nsan/intercep_strsep.cpp b/compiler-rt/test/nsan/intercep_strsep.cpp
new file mode 100644
index 00000000000000..4ae676fe20dcb1
--- /dev/null
+++ b/compiler-rt/test/nsan/intercep_strsep.cpp
@@ -0,0 +1,24 @@
+// RUN: %clangxx_nsan -O2 %s -o %t
+// RUN: %run %t 2>&1 | FileCheck %s
+
+#include <cstring>
+#include <iostream>
+
+extern "C" void __nsan_dump_shadow_mem(const char *addr, size_t size_bytes,
+                                       size_t bytes_per_line, size_t reserved);
+
+int main() {
+  // Define a C-style string with commas as delimiters
+  char input[] = "apple,banana,cherry,date";
+  char *token;
+  char *rest = input; // Pointer to keep track of the rest of the string
+
+  // Tokenize the string using strsep
+  while ((token = strsep(&rest, ",")) != NULL) {
+    std::cout << token << std::endl;
+  }
+
+  __nsan_dump_shadow_mem(&input[5], 1, 1, 0);
+  // CHECK: 0x{{[a-f0-9]*}}:    _
+  return 0;
+}