[Strings] Avoid mishandling unicode in StringConcat (#6411)

2 files changed, 32 insertions, 1 deletions

diff --git a/src/wasm-interpreter.h b/src/wasm-interpreter.h
index af01865f4..36542e6ed 100644
--- a/src/wasm-interpreter.h
+++ b/src/wasm-interpreter.h
@@ -1959,6 +1959,11 @@ public:
     if (!leftData || !rightData) {
       trap("null ref");
     }
+    // This is only correct if all the bytes in the left operand correspond
+    // to single unicode code points.
+    if (hasNonAsciiUpTo(leftData->values)) {
+      return Flow(NONCONSTANT_FLOW);
+    }
 
     Literals contents;
     contents.reserve(leftData->values.size() + rightData->values.size());
diff --git a/test/lit/passes/precompute-strings.wast b/test/lit/passes/precompute-strings.wast
index f70430648..c9facfe36 100644
--- a/test/lit/passes/precompute-strings.wast
+++ b/test/lit/passes/precompute-strings.wast
@@ -1,11 +1,21 @@
-;; NOTE: Assertions have been generated by update_lit_checks.py and should not be edited.
+;; NOTE: Assertions have been generated by update_lit_checks.py --all-items and should not be edited.
 
 ;; RUN: wasm-opt %s --precompute -all -S -o - | filecheck %s
 
 (module
+ ;; CHECK:      (type $0 (func (result i32)))
+
+ ;; CHECK:      (type $1 (func (result (ref string))))
+
  ;; CHECK:      (type $array16 (array (mut i16)))
  (type $array16 (array (mut i16)))
 
+ ;; CHECK:      (export "get_codepoint-bad" (func $get_codepoint-bad))
+
+ ;; CHECK:      (export "slice" (func $slice))
+
+ ;; CHECK:      (export "slice-bad" (func $slice-bad))
+
  ;; CHECK:      (func $eq-no (type $0) (result i32)
  ;; CHECK-NEXT:  (i32.const 0)
  ;; CHECK-NEXT: )
@@ -36,6 +46,22 @@
   )
  )
 
+ ;; CHECK:      (func $concat-bad (type $0) (result i32)
+ ;; CHECK-NEXT:  (string.eq
+ ;; CHECK-NEXT:   (string.concat
+ ;; CHECK-NEXT:    (string.const "a\f0")
+ ;; CHECK-NEXT:    (string.const "b")
+ ;; CHECK-NEXT:   )
+ ;; CHECK-NEXT:   (string.const "a\f0b")
+ ;; CHECK-NEXT:  )
+ ;; CHECK-NEXT: )
+ (func $concat-bad (result i32)
+  (string.eq
+   (string.concat (string.const "a\F0") (string.const "b"))
+   (string.const "a\F0b")
+  )
+ )
+
  ;; CHECK:      (func $length (type $0) (result i32)
  ;; CHECK-NEXT:  (i32.const 7)
  ;; CHECK-NEXT: )