Verify subject length before passing subject pointer to pcre2_match_8

src/bytes.rs

@@ -1259,6 +1259,19 @@ mod tests {
         assert!(re.is_match(b("a")).unwrap());
     }
 
+    // This tests a regression caused a segfault in the pcre2 library
+    // https://github.com/BurntSushi/rust-pcre2/issues/10
+    #[test]
+    fn jit_test_lazy_alloc_subject() {
+        let subject: Vec<u8> = vec![];
+
+        let re = RegexBuilder::new()
+            .jit_if_available(true)
+            .build(r"xxxx|xxxx|xxxx")
+            .unwrap();
+        assert!(!re.is_match(&subject).unwrap());
+    }
+
     #[test]
     fn utf_with_invalid_data() {
         let re = RegexBuilder::new()

src/ffi.rs

@@ -402,10 +402,20 @@ impl MatchData {
     pub unsafe fn find(
         &mut self,
         code: &Code,
-        subject: &[u8],
+        mut subject: &[u8],
         start: usize,
         options: u32,
     ) -> Result<bool, Error> {
+        // When the subject is empty, we use an empty slice
+        // with a known valid pointer. Otherwise, slices derived
+        // from, e.g., an empty `Vec<u8>` may not have a valid
+        // pointer, since creating an empty `Vec` is guaranteed
+        // to not allocate.
+        const EMPTY: &[u8] = &[];
+        if subject.is_empty() {
+            subject = EMPTY;
+        }
+
         let rc = pcre2_match_8(
             code.as_ptr(),
             subject.as_ptr(),